A Study of Sino-Korean Phonology: Its Origin, Adaptation and Layers 1315279797, 9781315279794

Table of contents :
Cover
Title
Copyright
Contents
Acknowledgments
Illustrations
Abbreviations
Conventions
Timeline of Chinese and Korean history
Preface
1 Introduction
1.1 Statement of purpose
1.2 Periodization of Chinese, Korean and Sino-Korean
1.2.1 Periodization of Chinese
1.2.2 Periodization of Korean
1.2.3 Periodization of Sino-Korean
1.3 Middle Chinese and Middle Korean phonology
1.3.1 Middle Chinese phonology
1.3.2 Middle Korean phonology
1.4 Literature review
1.4.1 Wu dialect origin hypothesis
1.4.2 Qieyun origin hypothesis
1.4.3 Song Dynasty (Kaifeng dialect) origin hypothesis
1.4.4 Tang Dynasty (Chang’an dialect) origin hypothesis
1.4.5 Old Chinese origin hypothesis
1.4.6 Summary of previous studies
1.5 Research methodology
1.6 Research data
1.6.1 Korean data
1.6.2 Chinese data
1.7 Research structure and organization
2 Mandarin loanwords in Modern Korean
2.1 Mandarin phonology
2.1.1 Mandarin consonants
2.1.2 Mandarin vowels
2.1.3 Mandarin syllables
2.1.4 Mandarin tones
2.2 Korean phonology
2.2.1 Korean consonants
2.2.2 Korean vowels
2.2.3 Korean syllables
2.2.4 Korean vowel length and pitch
2.3 Korean transcription systems for Mandarin loanwords
2.3.1 South Korean government transcription
2.3.2 Choe-Kim (1985)
2.3.3 Eom (1996)
2.3.4 Eom (2002)
2.4 Loanword phonology
2.4.1 Perception approach
2.4.2 Phonology approach
2.4.3 Phonetics-phonology approach
2.5 Discussion
2.5.1 Reflection on perception and phonology approaches
2.5.2 Influence of orthography
2.5.3 Role of the borrower
2.5.4 Summary
3 Initial system of Middle Sino-Korean
3.1 Group bang 幫 (bilabials)
3.2 Group fei 非 (labiodentals)
3.3 Group jian 見 (velars)
3.4 Group ying 影 (gutturals)
3.5 Group duan 端 (alveolar stops, laterals and nasals)
3.6 Group zhi 知 (retroflex stops and nasals)
3.7 Group jing 精 (alveolar fricatives and affricates)
3.8 Group zhuang 莊 (retroflex fricatives and affricates)
3.9 Group zhang 章 (palatal fricatives and affricates)
3.10 Summary
4 Final system of Middle Sino-Korean
4.1 Guo 果 rhyme group
4.2 Jia 假 rhyme group
4.3 Yu 遇 rhyme group
4.3.1 Mu 模 rhyme
4.3.2 Yu 魚 rhyme
4.3.3 Yu 虞 rhyme
4.3.4 Summary
4.4 Xie 蟹 rhyme group
4.4.1 Hai 咍 rhyme
4.4.2 Hui 灰 rhyme
4.4.3 Tai 泰 rhyme
4.4.4 Jie 皆 rhyme
4.4.5 Jia 佳 rhyme
4.4.6 Guai 夬 rhyme
4.4.7 Ji 祭 rhyme
4.4.8 Fei 廢 rhyme
4.4.9 Qi 齊 rhyme
4.4.10 Summary
4.5 Zhi 止 rhyme group
4.5.1 Zhi 支 rhyme
4.5.2 Zhi 脂 rhyme
4.5.3 Zhi 之 rhyme
4.5.4 Wei 微 rhyme
4.5.5 Summary
4.6 Xiao 效 rhyme group
4.6.1 Hao 豪 rhyme
4.6.2 Yao 肴 rhyme
4.6.3 Xiao 宵 rhyme
4.6.4 Xiao 蕭 rhyme
4.6.5 Summary
4.7 Liu 流 rhyme group
4.7.1 Hou 侯 rhyme
4.7.2 You 尤 rhyme
4.7.3 You 幽 rhyme
4.7.4 Summary
4.8 Xian 咸 rhyme group
4.8.1 Tan 覃 rhyme
4.8.2 Tan 談 rhyme
4.8.3 Xian 咸 rhyme
4.8.4 Xian 銜 rhyme
4.8.5 Yan 鹽 rhyme
4.8.6 Yan 嚴 rhyme
4.8.7 Tian 添 rhyme
4.8.8 Fan 凡 rhyme
4.8.9 Summary
4.9 Shen 深 rhyme group
4.9.1 Qin 侵 rhyme
4.9.2 Ji 緝 rhyme
4.9.3 Summary
4.10 Shan 山 rhyme group
4.10.1 Han 寒 rhyme and huan 桓 rhyme
4.10.2 Shan 山 rhyme
4.10.3 Shan 刪 rhyme
4.10.4 Xian 仙 rhyme
4.10.5 Xue 薛 rhyme
4.10.6 Yuan 元 rhyme
4.10.7 Xian 先 rhyme
4.10.8 Summary
4.11 Zhen 臻 rhyme group
4.11.1 Hen 痕 rhyme
4.11.2 Zhen 真 rhyme
4.11.3 Zhi 質 rhyme
4.11.4 Zhen 臻 rhyme
4.11.5 Yin 殷 rhyme
4.11.6 Hun 魂 rhyme
4.11.7 Zhun 諄 rhyme
4.11.8 Wen 文 rhyme
4.11.9 Summary
4.12 Dang 宕 rhyme group
4.12.1 Tang 唐 rhyme
4.12.2 Yang 陽 rhyme
4.12.3 Summary
4.13 Jiang 江 rhyme group
4.13.1 Jiang 江 rhyme
4.13.2 Summary
4.14 Zeng 曾 rhyme group
4.14.1 Deng 登 rhyme
4.14.2 Zheng 蒸 rhyme
4.14.3 Summary
4.15 Geng 梗 rhyme group
4.15.1 Geng 庚 rhyme (div. 2)
4.15.2 Geng 庚 rhyme (div. 3)
4.15.3 Geng 耕 rhyme
4.15.4 Qing 清 rhyme
4.15.5 Qing 青 rhyme
4.15.6 Summary
4.16 Tong 通 rhyme group
4.16.1 Dong 東 rhyme
4.16.2 Dong 冬 rhyme
4.16.3 Zhong 鍾 rhyme
4.16.4 Summary
4.17 Summary
4.17.1 Kaikou first division
4.17.2 Hekou first division
4.17.3 Kaikou second division
4.17.4 Hekou second division
4.17.5 Kaikou third division
4.17.6 Hekou third division
4.17.7 Kaikou fourth division
4.17.8 Hekou fourth division
5 Tonal system of Middle Sino-Korean
5.1 Tones in HMJH ( Yesan mungo edition)
5.2 Tone notation in other editions of HMJH
5.3 Reasons for erroneous tone notation in HMJH
5.4 Tone representation in other documents
6 Investigation into the origin of Sino-Korean
6.1 Initials
6.1.1 The bang 幫 initial group and the fei 非 initial group
6.1.2 The yun 云 initial
6.1.3 The zhi 知 initial group
6.1.4 The chan 禪 initial and the chuan 船 initial
6.2 Finals
6.2.1 Chongyun 重韻 (double rhymes)
6.2.2 The -j glide of the fourth division rhymes
6.2.3 The [ʌ] reading in the zhi 止 rhyme group
6.2.4 -l coda
7 Conclusion
7.1 Origin of Sino-Korean
7.2 Adaptation layers of Sino-Korean
7.3 Process of adaptation of Sino-Korean
7.3.1 Adaptation of initials
7.3.2 Adaptation of finals
7.3.3 Factors involved in adaptation
7.4 Irregular readings in Sino-Korean
7.5 Future work
References
Index

Citation preview

A Study of Sino-Korean Phonology

The term Sino-Korean may refer to either the phonological system or vocabulary in Korean that is of Chinese origin. Along with the borrowing of Chinese characters, the Chinese readings of characters must also have been transmitted into Korean. A Study of Sino-Korean Phonology aims to contribute to the field of Sino-Korean phonology by re-examining the origin and layers of Sino-Korean pronunciations from a loanword phonology perspective. The central issues of this book include an ongoing discussion on the questions of which Chinese dialect Sino-Korean is based on and how the source form in Chinese was adapted into Korean. Last is an in-depth analysis of the layers of Sino-Korean. Youyong Qian is Assistant Researcher at the Institute of Linguistics, Chinese Academy of Social Sciences. He received his PhD in Chinese Linguistics from the University of Wisconsin-Madison in 2015 and MA in Chinese Linguistics from Hanyang University, Seoul, South Korea, in 2010. His research interests include theoretical linguistics, phonology, Chinese historical phonology, Korean historical phonology and second language acquisition.

Routledge Studies in Chinese Linguistics Series editor: Hongming Zhang

Titles in the series: Dimensions of Variation in Written Chinese Zheng-Sheng Zhang Mandarin Chinese Words and Parts of Speech: Corpus-based Foundational Studies Huang Chu-Ren, Keh-Jiann Chen and Shu-Kai Hsieh Partition and Quantity Numeral Classifiers, Measurement, and Partitive Constructions in Mandarin Chinese (forthcoming) Jing Jin A Study of Sino-Korean Phonology: Its Origin, Adaptation and Layers Youyong Qian Syntax-Phonology Interface: Argumentation from Tone Sandhi in Chinese Dialects Hongming Zhang The Acquisition of Anaphora in Child Mandarin (forthcoming) Ruya Li The Interfaces of Chinese Syntax with Semantics and Pragmatics Yicheng Wu The Processing, Learning, and Teaching of Chinese Characters (forthcoming) Yun Yao The Semantics of Chinese Classifiers and Linguistic Relativity Song Jiang Further information on this series can be found at www.routledge.com/languages/ series/RSICL

A Study of Sino-Korean Phonology Its Origin, Adaptation and Layers Youyong Qian

First published 2018 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN and by Routledge 711 Third Avenue, New York, NY 10017 Routledge is an imprint of the Taylor & Francis Group, an informa business © 2018 Youyong Qian The right of Youyong Qian to be identified as the author of this work has been asserted by him in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging-in-Publication Data A catalog record for this book has been requested ISBN: 978-1-138-24164-0 (hbk) ISBN: 978-1-315-27981-7 (ebk) Typeset in Times New Roman by Apex CoVantage, LLC

Contents

Acknowledgmentsxi Illustrationsxii Abbreviationsxvii Conventionsxix Timeline of Chinese and Korean historyxx Prefacexxi 1 Introduction 1.1  Statement of purpose  1 1.2  Periodization of Chinese, Korean and Sino-Korean  2 1.2.1  Periodization of Chinese  2 1.2.2  Periodization of Korean  3 1.2.3  Periodization of Sino-Korean  4 1.3  Middle Chinese and Middle Korean phonology  6 1.3.1  Middle Chinese phonology  6 1.3.2  Middle Korean phonology  8 1.4 Literature review 10 1.4.1  Wu dialect origin hypothesis  10 1.4.2  Qieyun origin hypothesis  10 1.4.3  Song Dynasty (Kaifeng dialect) origin hypothesis  11 1.4.4  Tang Dynasty (Chang’an dialect) origin hypothesis  11 1.4.5  Old Chinese origin hypothesis  13 1.4.6  Summary of previous studies  13 1.5 Research methodology 16 1.6 Research data 17 1.6.1 Korean data 17 1.6.2 Chinese data 19 1.7  Research structure and organization  22

1

vi  Contents 2

Mandarin loanwords in Modern Korean 2.1 Mandarin phonology 24 2.1.1 Mandarin consonants 24 2.1.2 Mandarin vowels 26 2.1.3 Mandarin syllables 29 2.1.4 Mandarin tones 30 2.2 Korean phonology 31 2.2.1 Korean consonants 31 2.2.2 Korean vowels 32 2.2.3 Korean syllables 34 2.2.4  Korean vowel length and pitch  35 2.3  Korean transcription systems for Mandarin loanwords  35 2.3.1  South Korean government transcription  35 2.3.2 Choe-Kim (1985) 46 2.3.3 Eom (1996) 47 2.3.4 Eom (2002) 48 2.4 Loanword phonology 49 2.4.1 Perception approach 50 2.4.2 Phonology approach 51 2.4.3 Phonetics-phonology approach 52 2.5 Discussion 53 2.5.1  Reflection on perception and phonology approaches 53 2.5.2  Influence of orthography  54 2.5.3  Role of the borrower  54 2.5.4 Summary 56

24

3

Initial system of Middle Sino-Korean 3.1  Group bang 幫 (bilabials)  57 3.2  Group fei 非 (labiodentals)  62 3.3  Group jian 見 (velars)  63 3.4  Group ying 影 (gutturals)  63 3.5  Group duan 端 (alveolar stops, laterals and nasals)  79 3.6  Group zhi 知 (retroflex stops and nasals)  79 3.7  Group jing 精 (alveolar fricatives and affricates)  88 3.8  Group zhuang 莊 (retroflex fricatives and affricates)  94 3.9  Group zhang 章 (palatal fricatives and affricates)  94 3.10 Summary 104

57

4

Final system of Middle Sino-Korean 4.1  Guo 果 rhyme group  107

107

Contents vii 4.2  Jia 假 rhyme group  109 4.3  Yu 遇 rhyme group  111 4.3.1 Mu 模 rhyme  111 4.3.2 Yu 魚 rhyme  112 4.3.3 Yu 虞 rhyme  114 4.3.4 Summary 115 4.4  Xie 蟹 rhyme group  115 4.4.1 Hai 咍 rhyme  115 4.4.2 Hui 灰 rhyme  116 4.4.3 Tai 泰 rhyme  117 4.4.4 Jie 皆 rhyme  118 4.4.5 Jia 佳 rhyme  119 4.4.6 Guai 夬 rhyme  121 4.4.7 Ji 祭 rhyme  121 4.4.8 Fei 廢 rhyme  122 4.4.9 Qi 齊 rhyme  123 4.4.10 Summary 125 4.5  Zhi 止 rhyme group  125 4.5.1 Zhi 支 rhyme  125 4.5.2 Zhi 脂 rhyme  128 4.5.3 Zhi 之 rhyme  131 4.5.4 Wei 微 rhyme  132 4.5.5 Summary 133 4.6  Xiao 效 rhyme group  133 4.6.1 Hao 豪 rhyme  133 4.6.2 Yao 肴 rhyme  135 4.6.3 Xiao 宵 rhyme  136 4.6.4 Xiao 蕭 rhyme  136 4.6.5 Summary 137 4.7  Liu 流 rhyme group  137 4.7.1 Hou 侯 rhyme  137 4.7.2 You 尤 rhyme  138 4.7.3 You 幽 rhyme  140 4.7.4 Summary 140 4.8  Xian 咸 rhyme group  140 4.8.1 Tan 覃 rhyme  140 4.8.2 Tan 談 rhyme  141 4.8.3 Xian 咸 rhyme  142 4.8.4 Xian 銜 rhyme  142 4.8.5 Yan 鹽 rhyme  143 4.8.6 Yan 嚴 rhyme  144

viii  Contents 4.8.7 Tian 添 rhyme  145 4.8.8 Fan 凡 rhyme  145 4.8.9 Summary 146 4.9  Shen 深 rhyme group  146 4.9.1 Qin 侵 rhyme  146 4.9.2 Ji 緝 rhyme  148 4.9.3 Summary 148 4.10  Shan 山 rhyme group  149 4.10.1 Han 寒 rhyme and huan 桓 rhyme  149 4.10.2 Shan 山 rhyme  150 4.10.3 Shan 刪 rhyme  151 4.10.4 Xian 仙 rhyme  152 4.10.5 Xue 薛 rhyme  154 4.10.6 Yuan 元 rhyme  154 4.10.7 Xian 先 rhyme  156 4.10.8 Summary 157 4.11  Zhen 臻 rhyme group  157 4.11.1 Hen 痕 rhyme  157 4.11.2 Zhen 真 rhyme  158 4.11.3 Zhi 質 rhyme  160 4.11.4 Zhen 臻 rhyme  160 4.11.5 Yin 殷 rhyme  161 4.11.6 Hun 魂 rhyme  162 4.11.7 Zhun 諄 rhyme  163 4.11.8 Wen 文 rhyme  164 4.11.9 Summary 164 4.12  Dang 宕 rhyme group  164 4.12.1 Tang 唐 rhyme  164 4.12.2 Yang 陽 rhyme  166 4.12.3 Summary 168 4.13  Jiang 江 rhyme group  168 4.13.1 Jiang 江 rhyme  168 4.13.2 Summary 169 4.14  Zeng 曾 rhyme group  170 4.14.1 Deng 登 rhyme  170 4.14.2 Zheng 蒸 rhyme  171 4.14.3 Summary 173 4.15  Geng 梗 rhyme group  173 4.15.1 Geng 庚 rhyme (div. 2)  173 4.15.2 Geng 庚 rhyme (div. 3)  174 4.15.3 Geng 耕 rhyme  175

Contents ix 4.15.4 Qing 清 rhyme  176 4.15.5 Qing 青 rhyme  178 4.15.6 Summary 179 4.16  Tong 通 rhyme group  179 4.16.1 Dong 東 rhyme  179 4.16.2 Dong 冬 rhyme  181 4.16.3 Zhong 鍾 rhyme  181 4.16.4 Summary 183 4.17 Summary 183 4.17.1  Kaikou first division  183 4.17.2  Hekou first division  183 4.17.3  Kaikou second division  184 4.17.4  Hekou second division  184 4.17.5  Kaikou third division  184 4.17.6  Hekou third division  185 4.17.7  Kaikou fourth division  186 4.17.8  Hekou fourth division  186 5

Tonal system of Middle Sino-Korean 5.1  Tones in HMJH (  Yesan mungo edition)  188 5.2  Tone notation in other editions of HMJH 190 5.3  Reasons for erroneous tone notation in HMJH 191 5.4  Tone representation in other documents  191

188

6

Investigation into the origin of Sino-Korean 6.1 Initials 193 6.1.1  The bang 幫 initial group and the fei 非 initial group 193 6.1.2  The yun 云 initial  194 6.1.3  The zhi 知 initial group  194 6.1.4  The chan 禪 initial and the chuan 船 initial  195 6.2 Finals 196 6.2.1 Chongyun 重韻 (double rhymes)  197 6.2.2  The -j glide of the fourth division rhymes  201 6.2.3  The [ʌ] reading in the zhi 止 rhyme group  201 6.2.4 -l coda 204

193

7 Conclusion 7.1  Origin of Sino-Korean  209 7.2  Adaptation layers of Sino-Korean  211 7.3  Process of adaptation of Sino-Korean  218

209

x  Contents 7.3.1  Adaptation of initials  218 7.3.2  Adaptation of finals  219 7.3.3  Factors involved in adaptation  223 7.4  Irregular readings in Sino-Korean  225 7.5 Future work 225 References227 Index233

Acknowledgments

This book grew out of my doctoral dissertation at the University of WisconsinMadison. I would like to take this chance to acknowledge my PhD advisor Professor Hongming Zhang for imparting me with a deeper understanding of Chinese historical phonology and instructing me in how to construct stronger academic arguments and thinking more critically. I would also like to express my gratitude to Professor Ik-sang Eom, my academic advisor while I was a master’s student at Hanyang University, Seoul, South Korea, for introducing me to the field of SinoKorean Phonology. I am grateful to my PhD dissertation committee members: Professor Byung-jin Lim, Professor Marlys Macken, Professor Yafei Li and Professor Weihua Zhu for each of their invaluable suggestions and comments given at the defense of my PhD dissertation. I have incorporated many of their insightful suggestions into this book. My gratitude also goes to Professor Tae-Eun Kim for her advice and suggestions on Korean phonology and loanword phonology, which was particularly helpful to me in writing Chapter 2 of this book. I would like to thank my colleagues in the department of Chinese Language and Literature at Hanyang University, where I worked from September 2015 to August 2016, as well as my current colleagues at the Language Institute of the Chinese Academy of Social Sciences, Beijing, China, for their whole-hearted support and kind encouragement. I would also like to thank Camille Burns from the Routledge press for her detailed instruction on format and style. I wish to express much gratitude for her patience and assistance. My special thanks go to Lisa Salonen for her superb editorial support throughout the process of turning this manuscript into a book. Thanks also go to Wenbei Chen, Isaac Huben, Mengmeng Yang for proofreading the manuscript and providing many valuable suggestions. Finally, I thank my parents for their support and understanding. I would like to dedicate this book to them.

Illustrations

Figure 1.1

Syllable structure of Qieyun 8

Tables 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 1.10 1.11 1.12 1.13 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 2.10 2.11 2.12 2.13 2.14

Periodization of Chinese proposed by Karlgren (1926) 2 Revised periodization of Chinese 3 Periodization of Korean 3 Periodization of Sino-Korean proposed by Yu (1980) 5 Revised periodization of Sino-Korean 6 The initial system in Qieyun 6 Reconstructed final system in Qieyun 7 Consonant system of 15th-century Middle Korean 8 Monophthong system of 15th-century Middle Korean 9 Diphthong system of 15th-century Middle Korean 9 Hypotheses on the origin of Sino-Korean 14 Hypotheses on the layers of Sino-Korean 14 Potential reasons of non-mainstream readings in Sino-Korean 16 Examples of Mandarin loanwords in Korean 24 Inventory of Mandarin consonants 26 Inventory of Mandarin vowels (SR) 28 Inventory of Mandarin vowels (UR) 28 Traditional description of Mandarin syllable structure 29 Syllable types in Mandarin 29 Four tones in Mandarin 30 Inventory of Korean consonants 32 Monophthong system of Standard Korean 32 Monophthong system of the Seoul Korean 34 Diphthong system of the Seoul Korean 34 Syllable structure of Korean 35 Syllable types in Korean 35 Plosive consonants in Mandarin and Korean 36

Illustrations xiii 2.15 2.16 2.17 2.18 2.19 2.20 2.21 2.22 2.23 2.24 2.25 2.26 2.27 2.28 2.29 2.30 2.31 2.32 2.33 2.34 2.35 2.36 2.37 2.38 2.39 2.40 2.41 2.42 2.43 2.44 2.45 2.46 2.47 2.48 2.49 2.50 2.51 2.52 2.53 2.54 2.55 3.1 3.2 3.3 3.4

Korean transcription of Mandarin aspirated plosives Phonetic differences between Korean tense and lax stops Korean transcription of Mandarin unaspirated plosives Fricative consonants in Mandarin and Korean Korean transcription of Mandarin [f] Korean transcription of English [f] Korean transcription of Mandarin fricatives [s] and [ʂ] Korean transcription of Mandarin fricative [ɕ] Allophones of Korean fricative [s] and [s*] Korean transcription of Mandarin fricative [x] Affricates in Mandarin and Korean Korean transcription of Mandarin aspirated affricates Korean transcription of Mandarin unaspirated affricate [tɕ] Korean transcription of Mandarin unaspirated affricate [ts] and [tʂ] Approximants in Mandarin and Korean Korean transcription of Mandarin approximants Nasals in Mandarin and Korean Korean transcription of Mandarin nasals Korean government transcription of Mandarin consonants High vowels in Mandarin and Korean Korean transcription of Mandarin high vowels [i] and [y] Transcription of Mandarin high vowels [ɿ] and [ʅ] in Korean Mid vowels in Mandarin and Korean Korean transcription of Mandarin mid vowels [e] and [o] Korean transcription of Mandarin mid vowel [ɛ] Korean transcription of Mandarin mid vowels [ə] and [ɤ] Low vowels in Mandarin and Korean Korean transcription of Mandarin low vowel Korean transcription of Mandarin vowels Syllable correspondence between Mandarin and Korean (1) Syllable correspondence between Mandarin and Korean (2) Korean transcription of Mandarin [j] glide Korean transcription of Mandarin [wei] and [wən] Korean transcription of Mandarin [jou] (1) Korean transcription of Mandarin [jou] (2) Korean transcription of Mandarin tones Differences between Choe-Kim’s (1985) transcription and KGT Eom’s (1996) transcription of Mandarin consonants [p], [t] and [k] Eom’s (1996) transcription of Mandarin vowels Eom’s (2002) transcription of Mandarin vowels Influence of orthography on the transcription The group bang 幫 in EMSK The group fei 非 in EMSK The group jian 見 in EMSK The group ying 影 in EMSK

36 37 37 37 38 38 38 38 39 39 39 39 39 40 40 40 40 41 41 42 42 42 42 43 43 43 43 43 44 44 44 45 45 45 45 45 46 47 48 49 54 58 62 64 71

xiv  Illustrations 3.5 The group duan 端 in EMSK 80 3.6 The group zhi 知 in EMSK 85 3.7 The group jing 精 in EMSK 89 3.8 The group zhuang 莊 in EMSK 95 3.9 The group zhang 章 in EMSK 98 3.10 Mainstream representation of EMSK initials 104 3.11 Assignment of aspiration in HMJH 105 3.12 Analogy and assignment of aspiration in HMJH 106 3.13 Syllable preference and the assignment of aspiration in HMJH 106 4.1 The guo 果 rhyme group in EMSK 108 4.2 Mainstream representation of the guo 果 rhyme group in EMSK 109 4.3 The jia 假 rhyme group in EMSK 109 4.4 Mainstream representation of the jia 假 rhyme group in EMSK 111 4.5 The mu 模 rhyme in EMSK 111 4.6 The yu 魚 rhyme in EMSK 112 4.7 The yu 虞 rhyme in EMSK 114 4.8 Mainstream representation of the yu 遇 rhyme group in EMSK 115 4.9 The hai 咍 rhyme in EMSK 115 4.10 The hui 灰 rhyme in EMSK 116 4.11 The tai 泰 rhyme in EMSK 117 4.12 The jie 皆 rhyme in EMSK 118 4.13 The jia 佳 rhyme in EMSK 120 4.14 The guai 夬 rhyme in EMSK 121 4.15 The ji 祭 rhyme in EMSK 121 4.16 The fei 廢 rhyme in EMSK 122 4.17 The qi 齊 rhyme in EMSK 123 4.18 Mainstream representation of the xie 蟹 rhyme group in EMSK 125 4.19 The kaikou zhi 支 rhyme in EMSK 126 4.20 The hekou zhi 支 rhyme in EMSK 127 4.21 The kaikou zhi 脂 rhyme in EMSK 128 4.22 The hekou zhi 脂 rhyme in EMSK 130 4.23 The zhi 之 rhyme in EMSK 131 4.24 The wei 微 rhyme in EMSK 132 4.25 Mainstream representation of the zhi 止 rhyme group in EMSK 134 4.26 The hao 豪 rhyme in EMSK 134 4.27 The yao 肴 rhyme in EMSK 135 4.28 The xiao 宵 rhyme in EMSK 136 4.29 The xiao 蕭 rhyme in EMSK 137 4.30 Mainstream representation of the xiao 效 rhyme group in EMSK 137 4.31 The hou 侯 rhyme in EMSK 137 138 4.32 The you 尤 rhyme in EMSK 4.33 Mainstream representation of the liu 流 rhyme group in EMSK 140 4.34 The tan 覃 rhyme in EMSK 140 4.35 The tan 談 rhyme in EMSK 141 4.36 The xian 咸 rhyme in EMSK 142

Illustrations xv 4.37 The xian 銜 rhyme in EMSK 142 4.38 The yan 鹽 rhyme in EMSK 143 4.39 The yan 嚴 rhyme in EMSK 144 4.40 The tian 添 rhyme in EMSK 145 4.41 The fan 凡 rhyme in EMSK 145 4.42 Mainstream representation of the xian 咸 rhyme group in EMSK 146 4.43 The qin 侵 rhyme in EMSK 146 4.44 The ji 緝 rhyme in EMSK 148 4.45 Mainstream representation of the shen 深 rhyme group in EMSK 149 4.46 The han 寒 rhyme in EMSK 149 4.47 The shan 山 rhyme in EMSK 150 4.48 The shan 刪 rhyme in EMSK 151 4.49 The xian 仙 rhyme in EMSK 152 4.50 The xue 薛 rhyme in EMSK 154 4.51 The yuan 元 rhyme in EMSK 154 4.52 The xian 先 rhyme in EMSK 156 4.53 Mainstream representation of the shan 山 rhyme group in EMSK 158 4.54 The hen 痕 rhyme in EMSK 158 4.55 The zhen 真 rhyme in EMSK 159 4.56 The zhi 質 rhyme in EMSK 160 4.57 The zhen 臻 rhyme in EMSK 161 4.58 The yin 殷 rhyme in EMSK 161 4.59 The hun 魂 rhyme in EMSK 162 4.60 The zhun 諄 rhyme in EMSK 163 4.61 The wen 文 rhyme in EMSK 164 4.62 Mainstream representation of the zhen 臻 rhyme group in EMSK 165 4.63 The tang 唐 rhyme in EMSK 165 4.64 The yang 陽 rhyme in EMSK 166 4.65 Mainstream representation of the dang 宕 rhyme group in EMSK 168 4.66 The jiang 江 rhyme in EMSK 168 4.67 Mainstream representation of the jiang 江 rhyme group in EMSK 169 4.68 The deng 登 rhyme in EMSK 170 4.69 The zheng 蒸 rhyme in EMSK 171 4.70 Mainstream representation of the zeng 曾 rhyme group in EMSK 173 4.71 The geng 庚 rhyme (div. 2) in EMSK 173 4.72 The geng 庚 rhyme (div. 3) in EMSK 175 4.73 The geng 耕 rhyme in EMSK 175 4.74 The qing 清 rhyme in EMSK 177 4.75 The qing 青 rhyme in EMSK 178 4.76 Mainstream representation of the geng 庚 rhyme group in EMSK 179 4.77 The dong 東 rhyme in EMSK 179 4.78 The dong 冬 rhyme in EMSK 181 4.79 The zhong 鍾 rhyme in EMSK 181 4.80 Mainstream representation of the tong 通 rhyme group in EMSK 183 5.1 Tone transcription convention in HMJH 188

xvi  Illustrations Representation of MC tones in HMJH (YS edition) 189 MC sources of tone representation in HMJH 189 Tone notations in different editions of HMJH 190 Erroneous tone notation caused by analogy 191 Representation of the chan 禪 initial and the chuan 船 initial in HMJH 195 6.2 The reading of some cong 從 initial characters in HMJH 196 6.3 List of double rhymes 197 6.4 Distinction of double rhymes in EMSK 198 6.5 Representation of fourth division rhymes in EMSK 201 6.6 The [ʌ] reading of the zhi 止 rhyme group in EMSK 202 6.7 The fanqie spellings for the zhi 止 rhyme group characters in HLYY 202 6.8 The codas in the northwest dialect of MC 205 6.9 Comparison of transliterations given by Xuanzang and Hyecho 206 7.1 Comparison of EMSK and MC 210 7.2 Layers of Sino-Korean readings 212 7.3 Korean adaptation of Chinese initials 218 7.4 Reconstructed rhyme system of the source language of SK 220 7.5 Korean adaptation of Chinese vowels 221 7.6 Korean adaptation of Chinese [e] 222 7.7 Korean adaptation of Chinese [ə] 223 5.2 5.3 5.4 5.5 6.1

Abbreviations

Terms EMC EMSK LMC LMSK MC MSK OC OM SK

Early Middle Chinese Early Middle Sino-Korean Late Middle Chinese Late Middle Sino-Korean Middle Chinese Middle Sino-Korean Old Chinese Old Mandarin Sino-Korean

Books CJM Cheon ja mun 千字文 DD Donggyeong daehak 東京大學 edition (of Hunmong jahoe) DG Dongguk seorim 東國書林 edition (of Hunmong jahoe) DGJU Dongguk jeong’un 東國正韻 DH Daehag eonhae 大學諺解 GJ Gwangju 光州 edition (of Cheon ja mun) GJG Gyujang gak 奎章閣 edition (of Hunmong jahoe) GMH Gwangmun hoe 光文會 edition (of Hunmong jahoe) HD Hwadong jeongeum tongseog ungo 華東正音通釋韻考 HG Hyo’gyeong eonhae 孝經諺解 HLYY Yiqie jing yinyi 一切經音義 by Huilin 慧琳 HMJE Hunmin jeongeum 訓民正音 HMJH Hunmong jahoe 訓蒙字會 HWZY Hongwu zhengyun 洪武正韻 IG I’gyeong eonhae 易經諺解 JGG Jon’gyeong gak mungo 尊經閣文庫 edition (of Hunmong jahoe) JSYY Jinshu yinyi 晉書音義 JU Gyujang jeon’un 奎章全韻 JY Jung’yong eonhae 中庸諺解

xviii  Abbreviations MJ Maeng’ja eonhae 孟子諺解 NE Non’eo eonhae 論語諺解 NG Naegak mungo 內閣文庫 edition (of Hunmong jahoe) NGD No’geoldae eonhae 老乞大諺解 OP Jeon’un okpyeon 全韻玉篇 PTS Paktongsa eonhae 朴通事諺解 QYZZT Qieyun zhizhang tu 切韻指掌圖 SB Seokbong 石峰 edition (of Cheon ja mun) SE Seo’gyeong eonhae 書經諺解 SG Si’gyeong eonhae 詩經諺解 SGSG Samguk sagi 三國史記 SH Sohag eonhae 小學諺解 SSDZ Sisheng dengzi 四聲等子 SSTH Saseong tonghae 四聲通解 SU Sam’un seonghwi 三韻聲彙 SYCHT Shengyin changhe tu 聲音唱和圖 XYYY Yiqie jing yinyi 一切經音義 by Xuanying 玄應 YH Sinjeung yuhap 新增類合 YJ Yuk’jo beop’bo dan’gyeong eonhae 六祖法寶壇經諺解 YS Yesan mungo 叡山文庫 edition (of Hunmong jahoe) ZYYY Zhongyuan yinyun 中原音韻

Conventions

Modern Mandarin forms and Chinese historical phonology terms have been transcribed with Hanyu pinyin 漢語拼音 (often simply referred to as pinyin), which is the official Romanization system used in mainland China for Mandarin Chinese and which has increasingly become the international standard of Romanization for Mandarin Chinese in the past several decades. Korean words are given in the Revised Romanization of Korean, which is the official Korean language Romanization system in South Korea. However, names of Hangul alphabets are given in the Yale Romanization system for Korean, as this system uses a single consistent spelling for each alphabet irrespective of its context, unlike the Revised Romanization of Korean, which proposes different transcriptions for identical alphabetic spellings depending on their phonetic readings. Phonetic transcriptions are given in the International Phonetic Alphabet (hereafter abbreviated IPA). Following notation standards in Phonology literature, square brackets are used to refer to certain phonological features, such as [+voiced], or with IPA notation to specify specific phonetic values. The fanqie 反切 spelling of a Chinese character is placed in parentheses.

Timeline of Chinese and Korean history

China

Korea

. . .

. . . Gojoseon 古朝鮮 (?–AD 37)

Western Han 漢 (206 BC–AD 8)

Han Commanderies (108 BC–AD 313)

Eastern Han 漢 (25–220) Three Kingdoms (220–265) Western Jin 晉 (265–317) Eastern Jin 晉 (317–420) Southern and Northern Dynasties (420–589)

Silla 新羅 (57 BC– AD 668)

Goguryeo 高句麗 (37 BC– AD 668)

Baekje 百濟 (18 BC– AD 663)

Sui 隋 (581–618) Tang 唐 (618–907)

Unified Silla (668–935)

Five Dynasties (907–960) Northern Song 宋 (960–1127) Southern Song 宋 (1127–1279)

Goryeo 高麗 (918–1259)

Yuan 元 (1271–1368) Ming 明 (1368–1644) Qing 清 (1616–1911) . . .

Joseon 朝鮮 (1392–1910) . . .

Preface

This book will namely concentrate on topics in Sino-Korean phonology. In a broad sense, the term Sino-Korean refers to the phonological system and large body of vocabulary in Korean that are Sinoxenic — of Chinese origin. In this study, we will apply this term in its narrower sense to refer to the Korean reading of Chinese characters — which can be considered an important and salient piece of the phonological and orthographical interface of Korean’s Sinoxenic vocabulary. Although the exact time and process by which the Chinese characters were transmitted to Korea are still under much debate, however, given Korea’s geographical proximity to China, many with much certainty speculate that such occurred over a gradual, large-scale process of cultural exchange. Along with the borrowing of the Chinese character orthographical system, Chinese pronunciations or readings of Chinese characters were simultaneously adapted into the Korean language. Various hypotheses have been proposed regarding the origin and adaptation layers of Sino-Korean, including those proposed by Maspero (1920), Karlgren (1926), Arisaka (1936), Pak (1971), Kōno (1979), Lee (1994), Ito (2007), Zhang (2008) and Eom (2008, 2015), among others. So far, no consensus has been achieved with the origin and layers of Sino-Korean being an outstanding issue in the fields of Chinese and Korean historical linguistic research. This book aims to contribute specifically to the field of Sino-Korean phonology through re-­examining the origin and layers of Sino-Korean from a loanword phonology perspective. Central issues discussed in this book can be encapsulated in the following questions: (1) What is the exact origin of Sino-Korean? (2) How were Chinese source forms adapted into Korean? (3) How many layers exist in Sino-Korean? This book consists of seven chapters. Chapter 1 is the introduction of this book, where the purpose, methodology and source data of this study are discussed. In this chapter also is included the background information necessary for this research, such as the periodization of Chinese, Korean and Sino-Korean as well as a brief introduction to Middle Chinese phonology and Middle Korean phonology. Chapter 2 includes an account of the theory of loanword phonology and investigates mechanisms of loanword adaptation through a case study of Mandarin loanwords in Korean. The result of this case study shows that in order to clarify the adaptation patterns of Mandarin loanwords in Korean, both perception and phonology should be taken into account. This case study also demonstrates that

xxii  Preface borrowers often adopt different strategies, including different constraints and different rankings, or priority of constraints, in the process of borrowing. Chapter 3 investigates the syllabic initials of Early Middle Sino-Korean (EMSK), with special concentration given to the following issues: irregular aspiration, the merging of shetou 舌頭 and sheshang 舌上 and the merging of zhengchi 正齒 and chitou 齒頭, and the representation of xia 匣 initial and ri日 initial, among other topics. Chapter 3 closes with a summary of the mainstream representation of EMSK initials. Chapter 4 investigates into the syllabic finals of EMSK, with special attention given to the following issues: the [j] medial in the second division, the distinction of double rhymes, the distinction of yu 魚 and yu 虞 rhyme, the difference between third division dong 東 and zhong 鍾 rhyme, the merging of rhymes and the origin of -l coda. Chapter 4 closes with a summary of the mainstream representation of EMSK finals. Chapter 5 introduces the tonal system of EMSK recorded in historical documents and analyzes inconsistencies found in tone notation among different works. Chapter 5 also contains a comparison of tone notation in different editions of Hunmong jahoe 訓蒙字會 (1527) and offers an explanation for erroneous tone notation found in this document. Chapter 6 compares the sound system of EMSK and the phonological system in different works of Chinese, specifically, Qieyun 切韻 (601), Xuanying’s 玄 應 (?–661) Yiqie jing yinyi 一切經音義 (c. 650), He Chao’s 何超 Jinshu yinyi 晉書音義 (747), Huilin’s 慧琳 (737－820) Yiqie jing yinyi 一切經音義 (810), Shengyin changhe tu 聲音唱和圖 and Qieyun zhizhang tu 切韻指掌圖 (c. twelfth century). Chapter 7 is the conclusionary section of this book. This chapter summarizes the conclusions reached throughout this book and proposes a hypothesis about the origin and layers of Modern Sino-Korean pronunciations as well as how the initial and the final system in the source language were adapted into Korean. At the end of this chapter, we will outline tasks of future related research.

1 Introduction

1.1  Statement of purpose The main topic of investigation in this book is “the origin and layers of SinoKorean.” In a broad sense, the term Sino-Korean refers to the phonological system and large body of vocabulary in Korean that are Sinoxenic — of Chinese origin. In this study, we will apply this term in its narrower sense to refer to the Korean reading of Chinese characters — which can be considered an important piece of the phonological and orthographical interface of Korean’s Sinoxenic vocabulary. The importance of understanding Sino-Korean phonology in the study of Chinese historical phonology was first addressed in Karlgren’s work Etudes sur la phonologie chinoise (1926), in which Karlgren utilized Sino-Korean as the one of the language sources, in addition to Chinese dialects, for the reconstruction of the Middle Chinese phonological system. Since the pioneering work of Karlgren, Sino-Korean has been used as an important source of evidence in the study of some outstanding issues in the field of Chinese historical phonological research. Such issues include problems of chongniu 重紐 “rime doublets,” the phonetic value of the zhi 止 rhyme group, and other major questions. Before invention of the Korean alphabet Hangul in AD 1443, Chinese characters served as the primary orthographic system for writing Korean language. A thorough understanding of Sino-Korean is a prerequisite for the interpretation of Korean materials represented in Chinese characters — such as idu 吏讀 “clerk readings”; hyangchal 鄉札 “vernacular letters”; or gugyeol 口訣 “oral embellishment” — which are important data sources in the study of Korean historical phonology. In spite of the importance of Sino-Korean in Chinese and Korean historical phonology research, our current understanding of the nature of Sino-Korean is far from complete. Various hypotheses have been proposed regarding the origin of Sino-Korean, including Maspero (1920), Karlgren (1926), Arisaka (1936), Pak (1971), Kōno (1979), Lee (1994), Ito (2007), Zhang (2008) and Eom (2008, 2015), among other studies. However, despite a relatively large body of research, no consensus has been achieved, with the origin of Sino-Korean being an outstanding issue in the field of Chinese and Korean historical linguistic research. In this book, we will reexamine the origin and layers of Sino-Korean from a loanword phonology approach, with special attention given to the following questions: (1) What is

2  Introduction the exact origin of Sino-Korean? (2) How were Chinese source forms adapted into Korean? (3) How many layers exist in Sino-Korean?

1.2  Periodization of Chinese, Korean and Sino-Korean 1.2.1  Periodization of Chinese In order to facilitate a deeper discussion concerning the language contact between the Chinese languages and Korean language, this section will briefly address the periodization of Chinese, Korean and Sino-Korean. There are many different perspectives regarding periodization of Chinese. Given that our main concern is the phonological change in Chinese, we will discuss a periodization of Chinese proposal namely based on phonological criteria. Historically, building on major phonological changes in Chinese linguistic history, Karlgren (1926[1940]: 20–21)1 divided the history of Chinese into the six major periods, as shown in Table 1.1. Table 1.1 Periodization of Chinese proposed by Karlgren (1926) Proto-Chinese Archaic Chinese Ancient Chinese Middle Chinese Old Mandarin Modern Chinese

The period preceding earliest known literary documents The language of the Shijing 詩經 [Book of Odes] (c. 1000 BC) The language of the rhyme dictionary Qieyun 切韻 (AD 601) The language of the Song 宋 (960–1279) Dynasty rhyme tables The language of the Hongwu zhengyun 洪武正韻 (AD 1375) Present-day Mandarin

According to Norman (1988: 23), later scholars generally replaced the terms “Archaic Chinese” and “Ancient Chinese” with “Old Chinese” and “Middle Chinese,” respectively. In this book, we will adopt the updated terms “Old Chinese” and “Middle Chinese.” Next, in order to maintain the distinction of the earlier period of Ancient Chinese and Middle Chinese, Pulleyblank (1984) and Baxter (1992) divided the updated Middle Chinese into Early Middle Chinese (hereafter referred to as EMC) and Late Middle Chinese (hereafter LMC). They argued that such a distinction was still important given that the sound system reflected in Qieyun was significantly different to the language that would develop later, that is, the language reflected in the sound glosses represented in Yan Shigu’s 顏師 古 commentary to Hanshu 漢書 [Book of Han], Buddhist transcription practices in Huilin’s 慧琳 Yiqie jing yinyi 一切經音義 [The sound and the meaning of all scriptures], and poetic rhyming by unconventional poets. According to Pulleyblank (1984: 3, 60–63), EMC corresponds to the literary language in the sixth century, as reflected in Qieyun (601), and LMC corresponds to the Chang’an 長安 dialect between the seventh and eighth centuries, as reflected in the Song Dynasty rhyme tables, such as Yunjing 韻鏡 [Mirror of rhymes], Qiyinlue 七音略 [Summary of the seven sounds], among others. In order to observe the chronological differences between EMC and LMC, this study will follow Pulleyblank’s observations and refer to the period of Ancient Chinese in Karlgren’s periodization as

Introduction 3 EMC, and the period of Middle Chinese in Karlgren’s periodization as LMC. The revised periodization of Chinese is shown in Table 1.2. Table 1.2  Revised periodization of Chinese Proto-Chinese Old Chinese EMC LMC Old Mandarin Modern Chinese

The period preceding earliest known literary documents The language of the Shijing (c. 1000 BC) The language of the rhyme dictionary Qieyun (AD 601) The language of the Song (960–1279) Dynasty rhyme tables The language of the Hongwu zhengyun (AD 1375) Present-day Mandarin

1.2.2  Periodization of Korean The Korean language is generally considered to be part of the Ural-Altaic language family, although there still remains much disagreement on its genealogy. Table 1.3 is a periodization of Korean proposed by Lee (1972). Table 1.3  Periodization of Korean Old Korean Early Middle Korean Late Middle Korean Recent Korean Modern Korean

Korean up to the end of Unified Silla (935) The language of Goryeo 高麗 Dynasty (918–1392) From the beginning of the Joseon 朝鮮 Dynasty (1392) to the end of the 16th century From the beginning of the 17th century to the end of the 19th century From the beginning of the 20th century to the present day

In Lee’s proposal, Old Korean is a particularly long period because no clear evidence is available to divide the period any further. The primary materials used by linguists in the reconstruction of Old Korean include place names, personal names and titles of official positions as recorded in history books, idu 吏讀 “clerk readings,” hyangchal 鄉札 “vernacular letters” and Old Korean loanwords in Japanese. Due to the scarcity of Old Korean data and the difficulty in the interpretation of these data, reconstructing a homogeneous Old Korean phonological system is a daring task. The primary materials that reflect the EMK phonology are Jilin leishi 雞林類事 [A miscellaneous collection of things in Korea] and Hyangyak gugeup bang 鄉藥救急方 [Prescriptions for emergency treatment]. The book Jilin leishi, which contains a transcription of about 350 words in the Goryeo language, is one of the most important materials that account twelfth-century Korean language. The author of this book is Sun Mu 孫穆, who was sent from the Northern Song Dynasty to Goryeo (918–1392) as an emissary in 1103. Hyangyak gugeup bang, compiled in 1236, is the oldest extant medical book in Korea. This book contains the native Korean pronunciation of medicinal plants transcribed with Chinese characters. Since these two books use Chinese characters to transliterate native

4  Introduction Korean vocabulary, the interpretations of these data pose many challenges to historical linguists. Thanks to the development of Hangul, a phonetic symbol writing system which was invented in 1443 and officially promulgated in 1446, one can get a detailed idea of the phonology in Late Middle Korean. For this reason, in this study our discussion on ancient Korean is mainly based on the phonemic system of Late Middle Korean. A detailed discussion can be found later in this chapter. 1.2.3  Periodization of Sino-Korean Although the exact time and process by which the Chinese characters were transmitted to Korea are still under much debate; however, given Korea’s geographical proximity to China, many speculate that such occurred over a gradual, large-scale process of cultural exchange. This process may be said to have started around 108 BC when the Han 漢 Dynasty set up four commanderies — specifically, Lelang 樂浪, Lintun 臨屯, Xuantu 玄菟 and Zhenfan 真番 — in the northern regions of the Korean peninsula. This event marked a close political and cultural communication between China and Korea in ancient times. It also had the implication that the transmission of Chinese characters to Korea should have been no later than second century BC. Beginning in first century BC, the Korean peninsula transitioned into the Three Kingdoms Period, with the impending inception of Goguryeo 高句麗 (37 BC– AD 668), Baekje 百濟 (18 BC–AD 663) and Silla 新羅 (57 BC–AD 668). Goguryeo was positioned in the northern regions of the Korean peninsula. According to Samguk sagi (hereafter SGSG) 三國史記 [Historical records of the Three Kingdoms], the oldest extant Korean history book, Goguryeo began to use characters with the establishment of the kingdom.2 According to volume 18 in SGSG, in AD 372, King Fu Jian 苻堅 (337–385) of Former Qin 前秦 (350–394) sent a figure of Buddha and Buddhist sutras to Goguryeo,3 which accounted for a specific example in what should have been a much larger scale transmission of Buddhism into the Goguryeo kingdom from Northern China. Given that Goguryeo mainly communicated with Northern China by land route, it is believed that Goguryeo adopted the northern Chinese pronunciations of the time. Baekje was situated in the southern part of the Korean peninsula. According to SGSG, during the period of King Geunchogo 近肖古 (AD 346–375), the savant of the Grand Academy named Go Heung 高興 compiled the history book Seogi 書記 to make record of the history of Baekje.4 Baekje’s communication with China was mainly via passage of sea. According to SGSG, Baekje’s first tribute to China was made to the Eastern Jin Dynasty in AD 372,5 after which Baekje sent a series of approximately 14 tribute missions to the southern states of China (Eom 1991: 18). Silla was located at the southeastern part of the Korean peninsula. Since Goguryeo and Baekje isolated Silla from direct path to China, communication with China came much later for Silla than the other two kingdoms. After the sixth century, however, Silla defeated Baekje and gained control over the lower Han River region, which gave Silla a new channel for communication with Southern China via sea. According to Liangshu 梁書 [The history of Liang], Silla sent tributes

Introduction 5 to the Liang 梁 (502–557) Dynasty in Southern China starting in 521.6 After the establishment of the Tang 唐 (618–907) Dynasty, Silla was greatly influenced by the Tang Dynasty through political and cultural exchanges. According to the records in Jiu Tang shu 舊唐書 [The old history of the Tang], Silla sent tribute missions to Tang starting from the fourth year of the reign of Wude 武德 (621).7 In terms of cultural exchanges, Silla started to send groups of learners to Tang China to study Chinese classics in 640.8 Tongsa kangmok 東史綱目 [The outline of Korean history] states that hundreds of students from Silla studied abroad in Chang’an 長安, the capital of Tang, during the seventh and eighth centuries.9 In 757, King Kyongduk 景德 (742–765) of Unified Silla ordered to change the native Korean place names to Chinese style names. This implied that systematic and uniform reading of Chinese characters in SK had established at that time. Although Unified Silla was overturned in 918, the SK system in Unified Silla was generally adopted by later Goryeo and Joseon dynasties and lasted until Modern Korean. In summary, the borrowing of Chinese characters into Korean can be traced back to as early as second century BC, when the Han 漢 (206 BC–AD 220) Dynasty set up four commanderies in the northern part of the Korean peninsula. Along with the borrowing of Chinese characters, the Chinese pronunciations and readings of characters would have also necessarily been transmitted into the Korean language, and this brought about the earliest form of Sino-Korean. Yu (1980) proposed the following periodization of Sino-Korean. Table 1.4  Periodization of Sino-Korean proposed by Yu (1980) Old Sino-Korean

From the first borrowing of Chinese Characters to Silla’s unification in AD 668 Middle Sino-Korean From 668 to the Imjin 壬辰 war10 in 1592 Modern Sino-Korean From 1592 to the present

The unification of Korea in 668 implies the emergence of a common linguistic standard throughout the Korean peninsula. Consequently, Sino-Silla became the norm for the pronunciation and reading of Chinese characters among educated speakers in the kingdom. The later Imjin war deteriorated much Korean society, undoubtedly affecting the norm of Sino-Korean. Among affected aspects of Sino-Korean were that the triangle symbol used to write the ri 日 initial and the diacritic dots used to mark tones disappeared altogether after the Imjin war. In this study, we will adopt this periodization with a little adjustment in regard to the terms “Modern Sino-Korean” and “Middle Sino-Korean.” SK in the Modern period is by no means the same as it was in 1592 given sound changes that occurred in the Korean language since then. Specifically, the consonants [t] and [tʰ] palatalized as [tɕ] and [tɕʰ], respectively, when followed by a medial or nucleus [i] between the seventeenth and eighteenth centuries. Additionally, the diphthongs [ai] and [əi] were reduced to monophthongs [æ] and [e], respectively, between the eighteenth and nineteenth centuries. In this study, the term

6  Introduction “Modern Sino-Korean” will be used solely when referring to SK in the present day. This study considers Hangul matchumbeop tongiran 한글 맞춤법 통일안 [A proposition for the unification of Hangul orthography] issued by the Korean Language Society in 1933 as the turning point between Modern SK and Middle SK. The periods from 1592 to 1933 and from 668 to 1592 will be regarded as Late Middle Sino-Korean and Early Middle Sino-Korean, respectively. Our periodization can be expressed as follows: Table 1.5  Revised periodization of Sino-Korean Old Sino-Korean Early Middle Sino-Korean Late Middle Sino-Korean Modern Sino-Korean

From the first borrowing of Chinese Characters to Silla’s unification in AD 668 From 668 to the Imjin war in 1592 From 1592 to 1933 From 1933 to the present

1.3  Middle Chinese and Middle Korean phonology 1.3.1  Middle Chinese phonology In order to study the relationship between MC and SK, we will compare the Qieyun system and SK in this study. Qieyun has a pivotal position in the history of Chinese phonology. On one hand, it is a turning point between Old Chinese and Middle Chinese. On the other hand, the later development from Middle Chinese to Modern Chinese can be observed by comparing the Qieyun system and Chinese in later stages of linguistic development. The phonetic reconstruction of the Qieyun provided in this study is based on Pan and Zhang (2013). Table 1.6 is Pan and Zhang’s (2013) reconstruction of the initial system in Qieyun. Table 1.6  The initial system in Qieyun 幫 series 端 series 知 series 精 series 莊 series 章 series 牙 velars 喉 glottals

幫p 端t 知ʈ 精 ts 莊 tʂ 章 tɕ 見k 影ʔ

滂 pʰ 透 tʰ 徹 ʈʰ 清 tsʰ 初 tʂʰ 昌 tɕʰ 溪 kʰ 曉h

並b 定d 澄ɖ 從 dz 崇 dʐ 禪 dʑ 群g 匣ɦ

明m 泥n 娘ɳ 心s 生ʂ 書ɕ 疑ŋ

來l 邪z 船ʑ

日ȵ

以j

Before introducing the rhyme system of Qieyun, it is necessary to explain two important concepts in historical Chinese phonology. First, Qieyun categorized initials according to kaihe 開合 (the open- and closed-mouth distinction). Initials in MC can be divided into two types, namely normal and rounded consonants. Initials with normal consonants are referred to as kaikou 開口 (open mouth), the rounded initials Cʷ are referred to as hekou 合口 (closed mouth). There exists no kaikou and hekou contrast for labial initials. In historical Chinese phonology,

Introduction 7 hekou also include the following three cases: (1) rhymes in which the first vowel of the syllable final is the vowel [u], such as the dong 東 rhyme [uŋ]; (2) when the first vowel of the final is the vowel [o], an onglide [u̯] will be generated, and these rhymes are considered as hekou rhymes, such as the mu 模 rhyme [u̯ o]; (3) in the rhymes like wei 微, fei 廢, wen 文, yuan 元, han 寒, wu 物, yue 月 and he 曷, which all consist of a back or a central vowel and an ending with [+front] feature, usually an onglide [ṷ] would also be generated. Second, in Chinese historical phonology, the concept deng 等 (division, abbreviated div.) is used to describe the medial of a syllable. According to Pan and Zhang (2013: 4), div. 1 and div. 4 have no medial, div. 2 has -ɣ medial, div. 3/3A has -i medial and div. 3B has -ɣi medial. Pan and Zhang (2013: 4) proposed the following underlying form and surface form for the final system in Qieyun. Table 1.7  Reconstructed final system in Qieyun Phoneme Division

a 1

3(i)

Vowel

ɑ

a

-ø

歌

-ŋ

唐

-m

談

-n

寒

-i

泰

-u

豪

Coda

Coda

Vowel

3(i)

2(ɣ)

æ

ɛ

3A(i) 3B(ɣi) E

i 4

3A(i) 3B(ɣi) i

e

麻二

麻三

佳

支

陽

庚二

庚三

耕

清

青

嚴(凡)

銜

咸

鹽

添

侵

刪

山

仙

先

真(臻)

夬

皆

祭

齊

宵

蕭

廢

肴

脂

o 3(i)

2(ɣ)

3(i)

1

3(i)

ə

ɨ

ə

ɤ

o

o̟

魚

模

虞

侯

冬

鍾

東1

之

-ŋ

登

蒸

-m

覃

-n

痕

殷

-i

咍

微

江

元

幽 u

1

-ø

-u

2(ɣ)

ə

Phoneme Division

e

魂

1

3(i) u

東3

文

灰

尤

MC has four tones, namely ping 平 (level tone), shang 上 (rising tone), qu 去 (departing tone) and ru 入 (entering tone). There is no conclusion for the tonal value of each tone. A general accepted description of MC tones can be expressed

8  Introduction σ Initial Final [Medial]

Rhyme

Nucleus

[Ending]

[Onglide] Vowel Figure 1.1  Syllable structure of Qieyun

as follows: ping is a level tone; shang a rising tone; qu a falling tone; and ru an abrupt tone (Pan and Zhang 2013). According to Pan and Zhang (2013), the syllable structure in Qieyun system can be expressed as Figure 1.1. The ones given in square brackets are optional, as they represent the position of medials, glides and endings. Medial values include -i- and -ɣ-, and ending values include -i/-u/-m/-n/-ŋ/-p/-t/-k. The on-glide -ṷ has no phonetic value, and it is only a part of nucleus. 1.3.2  Middle Korean phonology An accurate record of ancient Korean’s phonological system has been available since the fifteenth century, when Hangul, an alphabetic writing system, was presented in Hunmin jeongeum (hereafter HMJE) 訓民正音 [Correct sounds to instruct people], which was created in 1443 and promulgated in 1446. Although it would be ideal to compare MC with Old Korean (Korean up to Unified Silla [AD 935]), the understanding of Old Korean is still quite limited given the scarcity of Old Korean phonological documentation. Therefore, this study will limit its discussion to the comparison of MC with the phonemic system of fifteenth century Middle Korean. Below is the consonant system of fifteenth-century Middle Korean proposed by Lee Ki-Moon (1972[2007]) with some revision. Table 1.8  Consonant system of 15th-century Middle Korean Place of articulation

Voiceless unaspirated

Voiceless aspirated

Tensed

labial dental palate-alveolar velar glottal

ㅂp ㄷt ㅈ ts, ㅅs ㄱk ㅎh

ㅍpʰ ㅌtʰ ㅊtsʰ ㅋkʰ

ᇪpp ᄯ tt ㅆss ᄭkk ᅘ hh

ᄫv

Nasal

Liquid

ㅁm ㄴn

ㄹl

ᅀz ㆁŋ ㅇø

The only revision in this study to Lee Ki-Moon’s (1972) proposal is the sound value of the ㅇ symbol. Lee (1972) concludes that the phonetic nature of the

Introduction 9 ㅇ symbol is voiced fricative [ɦ]. However, Pak (1971) claims that the ㅇ, as in Modern Korean, is used in the initial position to indicate a zero-initial syllable. Pak (1971: 306) states that according to Hunmin jeongeum haerye 訓民正音解例 [The explanation of the script in Hunmin jeongeum], the sound represented by this symbol is “light and void” and cannot be used in the final position. In this study, we will treat this symbol as a zero initial as Pak (1971) suggests. The system of monophthongs in the fifteenth-century Middle Korean, as shown in HMJE, is generally accepted as follows (Lee Ki-Moon 1972[2007]: 151): Table 1.9  Monophthong system of 15th-century Middle Korean High Mid Low

ㅣi

ㅡɨ ㅓə ㅏa

ㅜu ㅗo ᆞʌ

Diphthongs can be divided into on-glide diphthongs and off-glide diphthongs based on the order in which they appear. Middle Korean has both on-glide diphthongs and off-glide diphthongs. The following chart is taken from Ito (2007: 5). Table 1.10  Diphthong system of 15th-century Middle Korean ㅑia ㅘoa ㅐ ai ㅙ oai

ㅕ ㅝ ㅔ ㅖ ㅞ ㆋ

iə uə əi iəi uəi iuiəi

ㅛ io

ㅠiu

ㅚ oi

ㅟ ui ㆌ iui

ㆎ ʌi

ㅢ ɨi

ㅣ i[ij]

There are in total eight codas in Late Middle Korean, including -p, -t, -k, -m, -n, -ŋ, -s and -l. The coda clusters in the Korean orthography are single consonants in actual pronunciation (Lee 1972; Ito 2007). According to Kōno (1979), the syllable structure in Middle Korean is “IMVF,” in which “I” refers to syllabic initial; “M” medial; “V” nuclear vowel; and “F” final vowel. In HMJE, tones found in fifteenth-century Middle Korean were notated by placing dots to the left of a word: specifically, one dot for a qu tone, two dots for a shang tone and no dot for a ping tone. For example, 활 [hoal], 돌 [tol], 갈 [kal] and 붇 [put] have ping tone, shang tone, qu tone and ru tone, respectively. In Nogeoldae eonhae 老乞大諺解 [Hangul annotated Nogeoldae] (hereafter NGD) and Paktongsa eonhae 朴通事諺解 [Hangul annotated Paktongsa] (hereafter PTS), Choe Sejin 崔世珍 (1473–1542) states that “the tones in Korean can be divided into ping 平 (level) and ze 仄 (oblique). The tone that is mournful and calm belongs to ping tone, the tone that is stern and lifts belongs to shang tone, the tone that is straight and high belongs to qu tone and ru tone.”11 Based on this description, the phonetic value of tones in Middle Korean may be summarized as follows: ping tone as a level tone, shang tone a low-rising tone and qu/ru tones as high tones.

10  Introduction

1.4  Literature review Extensive literature has been published on the origin of Modern SK. The hypotheses proposed in previous studies can be classified into the following five categories. 1.4.1  Wu dialect origin hypothesis 1.4.1.1  Maspero (1920) Maspero (1920) examined the historical communication between Korea with China concentrated throughout a long span beginning with the Three Kingdoms period and ending with the Joseon Dynasty. Maspero (1920) further claimed that Goguryeo, which exhibited a close relationship with northern China, borrowed the Chinese pronunciation and character readings from northern Chinese, whereas Silla, which had frequent exchanges with southern China, probably borrowed its Chinese pronunciation and readings from the Wu 吳 dialect around the fifth century. Because Silla unified the Korean peninsula, Sino-Silla pronunciations won over Sino-Goguryeo pronunciations to become the mainstream of SK in the Joseon Dynasty. Maspero (1920) concluded that Modern SK must generally be based on the Wu dialect from the fifth century, although it also received some influence from Northern Chinese dialects. Maspero’s study, however, was rejected by Karlgren (1922: 4–6), who pointed out that Maspero’s hypothesis was merely built on historical probability. Maspero’s hypothesis was later reviewed in detail in Kōno (1979). Kōno (1979: 309) wrote that Maspero’s hypothesis was “filled with imagination.” However, Kōno (1979: 311) also conceded that “Maspero’s hypothesis had merit in its historical point of view, which is not seen in other studies.” 1.4.2  Qieyun origin hypothesis 1.4.2.1  Karlgren (1922, 1926) Karlgren was the first Western linguist who recognized the importance of SinoKorean in the reconstruction of Middle Chinese. After close comparison of SinoKorean and Sino-Japanese as well as many Chinese dialects, Karlgren (1922: 6, 1926: 76) rejected Maspero’s previous claims, concluding that Sino-Korean was essentially based on the Northern Chinese dialects of the Sui 隋 and early Tang 唐 dynasties, which was codified in the dictionary of Qieyun (601). 1.4.2.2  Pak Peong-cae (1971) Pak (1971) includes a comprehensive analysis on the initials, rhymes and tones of Middle SK. Pak (1971) proposed that the foundational stratum of SK is the northern Chinese dialect pronunciations of Sui and early Tang, as reflected in Qieyun, although some elements of the Wu dialect of the Southern Dynasties (420–589) can also be traced at the earliest layer, or stratum, of Sino-Korean.

Introduction 11 1.4.2.3  Zhang Weijia (2008) Zhang (2008) proposed that SK namely contains three layers, which are the “ancient layer, middle ancient layer, and near ancient layer.” The foundational layer of SK is a reflection of Middle Chinese, especially Early Middle Chinese. According to Zhang (2008), the [k] reading of the xia 匣 initial is a reflection of Old Chinese, and the palatalized reading of the zhi 知 initial (third division) and the duan 端 initial (fourth division) is the trace of the near ancient layer. In addition, Zhang (2008) attributes some features in SK, such as the distinction of yu 魚 and yu 虞, to the influence of the Wu dialect from the Southern Dynasties (420–589). 1.4.3  Song Dynasty (Kaifeng dialect) origin hypothesis 1.4.3.1  Arisaka (1936) Arisaka (1936) claimed that the origin of Sino-Korean is the tenth-century Kaifeng 開封 dialect of the Song Dynasty. According to Arisaka (1936), the following three prominent phonological features in SK indicate a close relationship between SK and the Chinese dialects of the Song Dynasty: First, in the zhi 止 rhyme group the vowel is represented as [ʌ] after sibilant initials. Second, Middle SK reflects yu魚 rhyme as [o] when the initial is from the zhuang 莊 initial group. Third, the -t coda in MC is represented as -l. Arisaka (1936) put forward that the coda -p, -l and -k in SK is a reflex of the Chinese coda system when the -p, -t and -k has weakened into -b, -r and -g. Arisaka’s hypothesis was rejected by Pak (1971) and Kōno (1979). Pak (1971: 282) criticized Arisaka’s theory, writing “Arisaka’s theory seems to have been drawn from a prejudiced deduction being based on limited facts concerning Sino-Korean, thereby losing a comprehensive view point.” Kōno (1979) points out that the fact that some features in SK being consistent with a certain period of Chinese does not necessarily means SK was borrowed from Chinese at that particular time. 1.4.4  Tang Dynasty (Chang’an dialect) origin hypothesis 1.4.4.1  Kōno (1979) Kōno’s study of SK can be differentiated with four distinct versions: dissertation (1961), serialized journal articles (1964–1967), monograph (1968) and the version in the Collected papers of Kōno Rokurō (1979). The introduction to Kōno’s study given here is based on Kōno (1979). Kōno (1979) claimed that in total four layers, roughly corresponding to the Wu dialect in the Southern dynasties, the Chang’an dialect in the Tang Dynasty, the Kaifeng dialect in the Song Dynasty and the Beijing dialect in the Ming 明 (1368–1644) Dynasty, can be found in SK. The main stratum of SK is based on the sound system in Chang’an in the Tang Dynasty, as reflected in the fanqie 反切12 in Huilin’s 慧琳 (737–820) Yiqie jing yinyi (810).

12  Introduction 1.4.4.2  Lee Haewoo (1994) Lee Haewoo (1994) provided an etymophonological comparison of Chinese dialects, Sino-Japanese and Sino-Korean. The degree of etymophonological affinity among these three language families is examined in detail in Lee (1994). As for the origin of SK, Lee (1994) concluded that “Sino-Korean originates from the Chang’an language of the northern Late Middle Chinese in the seventh and eighth centuries rather than the language of the Qieyun in the sixth century.” 1.4.4.3  Lee Yun-dong (1997) Lee (1997) conducted a close examination of the irregular readings in SK, proposing explanations for the irregular readings from different aspects, including analogy of phonetic radicals, the influence of Old Chinese or Old Mandarin, multiple fanqie notations, the influence of Chinese dialects or copy errors, among other factors. Lee (1997) concluded that the basis of SK is the Chang’an dialect in the Tang Dynasty. 1.4.4.4  Synn Asa (2006) Synn (2006) compared different stages of Sino-Korean, including Middle SK and Modern SK, with the sound system in different periods of Chinese, from Old Chinese to Modern Mandarin. Synn (2006) proposed that the starting point of SK is based on the Chinese from the third century to the Six Dynasties period in the sixth century, and the main stratum of SK is a reflection of the northern Chinese dialect in the eighth and ninth centuries, while the latest stratum is based on Chinese from the Northern Song Dynasty to the Yuan 元 (1271–1368) Dynasty. 1.4.4.5  Ito, Chiyuki (2007) Ito (2007) conducted a comprehensive research concerning the SK readings in 26 historical texts, which were published between the late fifteenth century and the late sixteenth century. Rejecting the Old Chinese origin hypothesis and the Qieyun origin hypothesis, Ito (2007) claimed that it is highly possible that the origin of SK was the Chang’an dialect in late Tang Dynasty, arguing that the pronunciation and reading of Chinese characters was probably transmitted into Korean by these Korean monks who went to China during late Tang Dynasty. 1.4.4.6  Eom, Ik-sang (2008, 2015) Eom (2008, 2015) examines the traces of Old Chinese, Middle Chinese, Old Mandarin and Modern Mandarin in the Modern Sino-Korean phonology system. According to Eom (2008), some readings in Modern SK such as [ta] 茶 “tea” and [kɯm] 金 “gold” are OC traces. Some seemingly native Korean words, such as [nʌ] 너 “you” and [na] 나 “I,” are also loanwords from OC. The matrix of

Introduction 13 SK, according to Eom (2008: 185), is the northern Chinese dialect in the sixth to eighth century. 1.4.4.7  Shin (2015) Shin (2015) proposed that since the Korean language has been continuously in contact with the Chinese language, Sino-Korean is a linguistic system consisting of multiple historical adaptation layers originating in different periods and regions of China. Shin (2015) demonstrated that Sino-Korean pronunciations not only reflect the features of Middle Chinese phonology but also comparable to Old Chinese and Post-Medieval Chinese phonology. 1.4.5  Old Chinese origin hypothesis 1.4.5.1  Nie Hongyin (1984) Nie (1984) argued that SK presents a faithful representation of the initial and rhyme system in Qieyun, but the SK readings of many characters, which cannot be explained from the perspective of Qieyun, are explainable from the perspective of Old Chinese. According to Nie (1984: 66), “the phonological system of SK is more close to Old Chinese than Qieyun. This is true as far as the issue of chongniu is concerned.” 1.4.5.2  Li Dechun (1985, 1994) Li (1985, 1994) proposed that many features in SK are a reflection of Old Chinese. According to Li (1985), such features in Middle SK as the fourth division yu 喻 initial being represented as an alveolar and the lack of distinction between alveolar stops and palatal stops, arose from the influence of Old Chinese. 1.4.5.3  Kang Sinhang (2003, 2010) Kang (2003, 2010) also traces the origin of SK to Old Chinese. Kang (2010) proposed that the origin of SK is a Chinese in which there was no distinction of alveolar stops (Kang using the Chinese phonological term shetou 舌头) with retroflex stops (sheshang 舌上), bilabial (zhongchun 重唇) and labial-dental (qingchun 輕唇), alveolar affricate/fricative (chitou 齒頭) and retroflex affricates and fricatives (zhengchi 正齒). Kang (2010: 98) stated that “the establishment of SK system was deeply influenced by the Chinese in the Weijin Nanbeichao 魏晉南北 朝 [Wei, Jin, North-South Dynasties] period (AD 220–589), when Buddhist classics were transmitted to the Korean peninsula.” 1.4.6  Summary of previous studies Based on the literature review above, the hypotheses on the origin of SK can be summarized as below.

14  Introduction Table 1.11  Hypotheses on the origin of Sino-Korean Proposal

The origin of SK

Kang (2003, 2010) Maspero (1920) Karlgren (1926) Pak (1971) Zhang (2008) Kōno (1979) Lee Yundong (1997) Eom (2008) Lee Haewoo (1994) Synn (2006) Ito (2007) Arisaka (1936)

Time

Source Language

Weijin Nanbeichao period (220 to 589) 5th century AD 600

Old Chinese

Sui and Tang dynasties (6–7th century) Early Middle Chinese Tang Dynasty (8–9th century) Tang Dynasty 6–8th century 7–8th century 8–9th century late Tang Dynasty Song Dynasty (10th century)

Wu dialect Northern Chinese (reflected in Qieyun) Northern Chinese (reflected in Qieyun) Chinese reflected in Qieyun Chang’an dialect (reflected in HLYY) Chang’an dialect Northern Chinese Chang’an dialect Northern Chinese Chang’an dialect Kaifeng dialect

It is widely believed that Sino-Korean pronunciations consist of different layers, but the number and the nature of layers are still under discussion, with no clear consensus (Pak 1971; Kōno 1979; Synn 2006; Zhang 2008; Shin 2015, etc.). Table 1.12  Hypotheses on the layers of SK Pak (1971)

Kōno (1979)

Synn (2006)

Zhang (2008)

Shin (2015)

Wu dialect (Southern Dynasties) Qieyun

Wu dialect (Southern Dynasties) Chang’an dialect (Tang Dynasty)

Chinese (3–6th century)

Ancient layer

Old Chinese

Kaifeng dialect (Song Dynasty)

Beijing dialect (Ming Dynasty)

Northern Chinese Middle ancient Middle Chinese dialect (8–9th layer century) Chinese Near ancient Post-Medieval (The Northern layer Chinese Song Dynasty– the Yuan Dynasty)

The controversy on the origin and layers of SK can be attributed to the following three reasons. The first reason is that some previous studies have tended to overlook that SK is a loanword system, rather than a Chinese language. For example, Karlgren (1954) represented SK readings as a supposed “foreign dialect,” placing it as

Introduction 15 a parallel sister branch to native Chinese dialects, such as Cantonese or South Min. Next, Karlgren used SK data as direct evidence in a comparative study of Middle Chinese phonology. However, the nature of SK is clearly much different to Chinese dialects. SK inevitably involves an unfaithful representation of Chinese sounds given that Chinese sounds in Korean passed through a filter of linguistic borrowing due to differences in the phonology in the two language systems. More to this fact, some previous studies, such as Kōno (1979), have noticed the existence of several unfaithful representations in SK. Specifically, Kōno (1979: 341) stated that the mismatch between the Qieyun system and SK could be explained by the phonological structure of Korean. However, the principle of the loanword adaptation in SK was not discussed whatsoever in Kōno (1979). Kōno (1979) proposed that SK was borrowed from the Chinese reflected in Huilin’s Yiqie jing yinyi, arguing that the merging of rhymes in Huilin’s fanqie 反切 is consistent with the rhyme system in SK. However, recent studies in loanword phonology have shown that the mismatch in the inventory of initials in two languages may cause two different initials to be represented as the same initial in the target language. These studies have also depicted loanword adaptation as being a more complicated process than previously assumed. For example, the inputs from the source language that do not exist in the target language will be replaced with the most similar element that exists in the target language. But the resulting level of similarity and whether similarity is actuated at the perception level or the phonemic level are significant and outstanding questions in loanword phonology. This issue, however, has not been discussed in previous studies concerning SK, and thus, we find that such needs to be investigated further in future studies of SK. Second, it is noteworthy that previous literature often presented very different explanations for the same phenomena or features in SK, bringing about much inconsistency in research. For example, the [k] reading of the xia 匣 initial in SK was considered as a reflection of OC in Zhang (2008), but the same was conversely considered in Pak (1971) and Kōno (1979) as a result of sound change caused by the analogy of the sound of the phonetic component or characters with the same phonetic radical. Similarly, the [o] for yu 魚 rhyme is argued to be borrowed feature from OC in Pak (1971), but conversely considered a borrowing from the Song Dynasty Chinese in Arisaka (1936) and Kōno (1979). The third reason involves the researcher’s understanding of Chinese historical phonology, and more specifically the distinction between the sound system of OC and MC, which differed from study to study. For example, Li (1994: 68) claimed the reason there exists no distinction between shetou (alveolar stop and nasal) and sheshang (retroflex stop and nasal) initials in SK is a result of influence from OC, arguing that one of the biggest differences between OC and MC lies in the fact that the sheshang initial group in MC does not exist in OC. However, no distinction between sheshang and shetou is viewed as one of the features in SK reflecting the EMC sound system in Zhang (2008: 5). According to Zhang (2008), the lack of the sheshang initial group is a phenomenon not limited to OC but rather remains a feature of Chinese until EMC.

16  Introduction

1.5  Research methodology This book attempts to resolve the controversy over the origin and the layers of Sino-Korean by stressing the following points. First, Sino-Korean phonology, i.e. the Korean readings of Chinese characters, involves the mapping between two phonological systems in the process of borrowing, and thus can be examined under the framework of loanword phonology; specifically, when a form in the source language cannot find its perfect match in the target language, an adjustment will be made. There are two types of adjustments. The first type is faithful adjustment, which includes the phonological processes of epenthesis, resyllabification and replacement. The second type is unfaithful adjustment, which can be summarized namely by the process of deletion. Which elements can be added, deleted or replaced is not a random selection; it is relatively difficult for us to state precisely the rule of adaptation in SK, but fortunately, the strategy of adaptation in Mandarin loanwords in Modern Korean could give us a hint on this issue. The Mandarin loanwords transmitted into Modern Korean — such as [u.lung.tɕʰa] 烏龍茶 “oolong tea”, [y.an] 元 “Chinese currency” — in some sense, can tell us the strategy of adaptation from Chinese to Korean. In order to have a better understanding of the adaptation strategy in SK, we will conduct a case study on the adaptation of Mandarin loanwords in Korean in this study. Second, for a certain feature in Sino-Korean, there are many potential origins. In order to avoid jumping to a quick conclusion, all the possible reasons listed in Table 1.13 will be examined in this study. Table 1.13 Potential reasons of non-mainstream readings in Sino-Korean ① ② ③ ④ ⑤ ⑥ ⑦

Sound change in the source language Analogy Different fanqie spellings Phonotactic constraints in Korean Adjustment in the process of adaptation Sound change in Korean Different source languages

It is necessary to emphasize that before claiming that a non-mainstream reading in SK is a reflection of a particular layer, we need to exclude factors one to five. Factor six “sound change in Korean” leads to the inner source layer, while the source inputs from Chinese in different periods create different outer source layers. Both the inner source layer and outer source layer exist in SK and should be distinguished clearly. Third, the main approach to investigating the origin of SK is through a close comparison between SK and Chinese in different periods. However, it should be noted that for each period, the sound system could vary from time to time and from place to place. So it is relatively vague to state merely that a certain feature in SK is a reflection of OC or MC. In order to have a more accurate comparison

Introduction 17 between SK and different stages and regional variations of Chinese, we will pay more attention to the sound system as reflected in particular historical documents, such as Qieyun (601), XYYY (c.650), JSYY (747), HLYY (810) etc., to set up a reference for time and place.

1.6  Research data 1.6.1  Korean data 1.6.1.1  Main data The earliest SK data with Hangul annotation that we have access to is from the Early Middle Sino-Korean (hereafter EMSK) period. The following three books discussed immediately below were compiled before the Imjin war (1592) and are the main data sources in this study. 1.6.1.1.1  HUNMONG JAHOE

Hunmong jahoe (hereafter HMJH) 訓蒙字會 [Collection of characters for training the unenlightened] is a character primer compiled in 1527 by Choe Sejin 崔 世珍 (1473–1542), who was an expert in Chinese language and who had long been an official interpreter in the Korean embassy in Beijing. HMJH is the earliest extant character dictionary in Korea which reflects the reading of SK in actual speech. Before HMJH, there was a rhyme book called Dongguk jeongun 東國正 韻 (hereafter DGJU), which was compiled in 1447. As a prescribed SK system, DGJU did not faithfully reflect character readings in Korean speech. Among other things, the phonetic notation in DGJU included voiced initial [b], [d] and [g], which do not exist in Korean. Several versions of HMJH have appeared since its first print. The earliest extant edition is the Yesan mungo 叡山文庫 edition (hereafter YS edition). The HMJH version investigated in this study is the YS edition published by the Institute of East Asian Studies in Dankook University in 1971. In this book besides the photocopied YS version, a photocopy of the Donggyeong daehak 東京大學 edition (hereafter DD edition) and a comprehensive introduction to HMJH are also included. HMJH gives both Sino-Korean and native Korean readings as well as definitions for 3,360 characters. The following seven characters appear twice in HMJH: tio 朝, tshiu/hoan 萑, tsia/tsiək 炙, kiu 葵, sa 沙, koan 觀, thiuk/iu 柚, and hʌiŋ/haŋ 行 appears three times. The reading given by Choe is based on actual pronunciation at that time. In some cases, Choe also provides the standard pronunciation when the popular reading is different from the standard pronunciation in his mind. For example, the notation for the character ai 埃 in HMJH is [ai], followed by a note saying that the standard reading is [ʌi]. This shows that the pronunciation of the character ai 埃 was [ai] in SK at that time, even though the standard pronunciation was [ʌi] in Choe’s mind.

18  Introduction 1.6.1.1.2  CHEON JA MUN

Cheon ja mun (hereafter CJM) 千字文 [Thousand character classic] is a Chinese character primer originally compiled by Zhou Xingsi 周興嗣 (470–521) in the Liang 梁 (AD 502–557) Dynasty. It consists of, in total, 1,000 Chinese characters, arranged into 250 four-line stanzas. The Cheon ja mun discussed here is the Hangul annotated edition, with the meaning and Korean pronunciation provided. The earliest extant edition is the Gwangju 光州 edition (hereafter GJ edition), which was published in 1575. No tone notation is provided in this edition. The most well-known edition in Korea is the Seokbong 石峰 edition (hereafter SB edition), which was published in 1583. Tone notation was provided in this edition, with circles drawn on the top left of characters to represent shang 上 tone and on the top right of characters to represent qu 去 tone. No tone notation was provided for ping 平 or ru 入 tones. The Cheon ja mun investigated in this study is the GJ edition published by the Institute of East Asian Studies in Dankook University in 1973. 1.6.1.1.3  SINJUNG YUHAP

Sinjung yuhap (hereafter YH) 新增類合 [An enlarged edition of Yuhap] is a Chinese character primer compiled by Yu Huichun 柳希春 (1513–1577). YH consists of two volumes and provides Sino-Korean readings as well as definitions for 3,000 characters. Tone notation was provided, with circles drawn on the top left of characters for shang 上 tone, and on the top right for qu 去 tone, and no tone notation was provided for ping 平 or ru 入 tones. The most famous three editions of YH are the Nason 羅孫 edition, the Ilsa 一蓑 edition and the Dongyang mungo 東洋文庫 edition. Among these three editions, the Nason 羅孫 edition is the most complete edition that we can access to today. The YH investigated in this study is the Nason edition published by the institute of East Asian studies in Dankook University in 1972. 1.6.1.2  Supplementary data In order to have a comprehensive understanding of the SK data, we will compare the main data source HMJH, CJM and YH with the data in other sources, which are listed below. It is worth noting that due to some difficulty to access ancient documents, the supplementary data in this study is mainly cited from the collected readings of SK included in Kōno (1979). In addition, the SK readings in DH and YJ as well as the reading in the JGG, NG, GJG and DG editions of HMJH provided in this book is cited from Ito (2007). 1.6.1.2.1  HWADONG JEONGEUM TONGSEOG UNGO

Hwadong jeongeum tongseog ungo 華東正音通釋韻考 (hereafter HD) was compiled in 1747 by Pak Seongwon 朴性源 (1697–1767) under the order of King Yeongjo 英祖 (1694–1776). The hwa 華 and dong 東 in the title of this book

Introduction 19 refers to China and Korea, respectively. For each Chinese character, the standard pronunciation in SK, “dongum 東音”, and the pronunciation in Chinese at that time, “hwaum 華音”, is provided in this book. At the top of each page, one can find the sokum 俗音 “popular reading” of some characters in SK. 1.6.1.2.2  SAMUN SEONGHWI

Sam’un seonghwi 三韻聲彙 (hereafter SU) was compiled by Hong Gyehui 洪啓 禧 (1703–1771) in 1751. Not following traditional Chinese rhyme books, which arrange characters according to the four tones, this book arranges characters according to SK readings’ Hangul spelling. Both Korean readings and Chinese readings are provided in this book. 1.6.1.2.3  GYUJANG JEONUN

Gyujang jeon’un 奎章全韻 (hereafter JU) was compiled by Lee Deok-mu 李 德懋 in 1792 and published in 1796 under the order of King Jeongjo. The book was formatted so that the four tones of SK were under four parallel sections on each one page. This book also provides both the Chinese reading (hwaum 華音) and Sino-Korean reading (dongum 東音). The appendix of this book includes an index called Okpyeon 玉篇. With a detailed introduction to the reading and definition of Chinese characters, this index can be viewed as one independent character dictionary, and it is traditionally called Jeonun okpyeon 全韻玉篇 (hereafter OP). 1.6.1.2.4  HANGUL ANNOTATION OF CLASSICS

Many projects were initiated to annotate the Buddhist and Confucius classics after the invention of Hangul. The result of these projects was a series of so-called eonhae 諺解 “Hangul annotation” works, which included but was not limited to the Hangul annotated version of dharmic treasure altar-sutra of the Sixth Patriarch 六祖法寶壇經 and the Hangul annotated version of Lunyu 論語 [The Analects of Confucius] etc. These materials are of particular importance because, as opposed to rhyme books which perhaps had the function of propagating a potential artificial prescriptive standard, the readings in these Hangul annotated classics were more faithful to the actual pronunciation. 1.6.2  Chinese data 1.6.2.1  Main data 1.6.2.1.1 QIEYUN

Qieyun 切韻, chiefly edited by Lu Fayan 陸法言 (AD 581–618), was compiled in the first year of the Renshou 仁壽 era (AD 601) of the Sui 隋 Dynasty. Disregarding disagreements on the nature of Qieyun, according to Pan Wuyun and

20  Introduction Zhang Hongming (2013), Qieyun mainly reflects the literary pronunciation of Nanjing 南京 and Luoyang 洛陽 around the seventh century. Much of Qieyun has been lost, with only fragments of it remaining. There were several expansions and revisions of Qieyun after the Tang Dynasty, of which the two most complete versions are Kanmiu buque qieyun 刊謬補缺切韻 [Corrected and supplemented Qieyun], which was compiled by Wang Renxu 王仁昫 (c. 656 – c.710) in the Tang Dynasty, and Dasong chongxiu guangyun 大宋重修廣韻 (usually simply referred to as Guangyun 廣韻) [The revised dictionary of rhymes of the Song Dynasty], which was compiled by Chen Pengnian 陳彭年 (AD 961–1017) in the Northern Song Dynasty. The Qieyun series of rhyme dictionaries are the central archival sources in the reconstruction of Middle Chinese phonology. 1.6.2.1.2 XUANYING’S YIQIE JING YINYI

According to Zhou Fagao (1948), Xuanying 玄應 (AD ?–661) was a monk in the Dazongchi 大總持 temple and the Dacien 大慈恩 temple in Chang’an 長 安 during the Tang Dynasty. He joined Xuanzang’s 玄奘 (602–664) Buddhist classics translation project in the nineteenth year of the reign of Zhenguan 貞 觀 (AD 645) and probably passed in the first year of the reign of Longshuo 龍朔 (AD 661). Xuanying’s Yiqie jing yinyi (abbr. XYYY) 一切經音義 [The sound and the meaning of all scriptures] was compiled around AD 650. The fanqie in XYYY reflects the sound system of the Chang’an dialect in the mid-seventh century. Our understanding of the phonological system of XYYY is mainly based on the work of Zhou Fagao (1948) and Wang Li (1982a). 1.6.2.1.3  JINSHU YINYI

Jinshu yinyi (hereafter JSYY) 晉書音義 [The sound and the meaning of Jinshu] was written by He Chao 何超 in the Tang Dynasty. The preface of this book was written by Yang Qixuan 杨齊宣 in the sixth year of the reign of Tianbao 天宝 (AD 747). In this preface Yang mentioned that He Chao was a scholar from the Eastern capital of Luoyang 洛陽. Therefore, the fanqie in this book is a reflection of the sound system of Luoyang in the mid-eighth century. Our understanding of the sound system in JSYY is based primarily on an investigation conducted by Shao Rongfen (1981). 1.6.2.1.4  HUILIN’S YIQIE JING YINYI

Huilin 慧琳 (AD 737－820) was a monk in the Ximing 西明 temple during the reigns of Zhenyuan 貞元 and Yuanhe 元和 in the Tang Dynasty. The Yiqie jing yinyi 一切經音義 [The sound and the meaning of all scriptures] (abbr. HLYY) compiled by Huilin has 100 volumes. According to Song gaoseng zhuan 宋高 僧传 [a collection of biographies of Buddhist monks] written by Zanning 赞宁 (919–1001), Huilin started to write this book in the fourth year of the reign of Zhenyuan 貞元 (788) and ends in the fifth year of the reign of Yuanhe 元和 (810).

Introduction 21 According to the preface in HLYY, the fanqie provided in this book is based on the Qin 秦 pronunciation, which refers to the sound system in the region around Chang’an. Therefore, the fanqie in HLYY is a reflection of the sound system in the Chang’an dialect in the late eighth century to early ninth century. Our analysis of HLYY in this study heavily relies on the work of Huang Cuibo (1931), who categorized the initial, final and tonal system of HLYY. 1.6.2.1.5  SHAO RONG’S SHENGYIN CHANGHE TU

Shengyin changhe tu 聲音唱和圖 [The graph of matching sounds] is a chart included in Huangji jingshi 皇極經世 [Book of supreme world ordering principles], a Daoist treatise written by philosopher Shao Yong 邵雍 (1011–1077) in the Northern Song Dynasty (960–1126). This chart is arranged according to the initial and rhyme system of Bianliang 汴梁 (i.e. Kaifeng 開封) and Luoyang in the eleventh century. The information in this chart serves as one of the major sources of evidence used to reconstruct the sound system of the Luoyang dialect in the Northern Song Dynasty in Zhou Zumo (1966). 1.6.2.1.6.  RHYME TABLES IN THE SONG DYNASTY

A rhyme table is a phonological table used to arrange Chinese characters according to their tones, initials and finals to indicate their pronunciations. The earliest rhyme tables are used to assist the understanding of the sound system in Guangyun, such as Yunjing 韻鏡 [Mirror of rhymes] and Qiyinlue 七音略 [Summary of the seven sounds]. However, later rhyme tables are made based on the sound system in the spoken language at time recorded, including Sisheng dengzi (abbr. SSDZ) 四聲等子 [A measurement criteria of sounds] and Qieyun zhizhangtu (abbr. QYZZT) 切韻指掌圖 [An easy guide to Qieyun]. There is no clear information about the author and the exact date of SSDZ. However, the information reflected in SSDZ shows that it was probably a work from the early Southern Song period. QYZZT is a rhyme book attributed to Sima Guang 司馬光 (1019–1086), a famous chancellor and scholar who lived in the Song Dynasty, although the true author of this book is uncertain. Sun Yi 孫奕, a Southern Song Dynasty scholar, cited QYZZT in his work, suggesting that QYZZT existed in the Southern Song. These two rhyme tables are more valuable than earlier rhyme tables in the sense of reflecting the sound system in the actual speech. 1.6.2.2  Supplementary data 1.6.2.2.1 TRANSLITERATION

Transliteration materials refers to the Chinese transcription of foreign words in ancient texts, particularly Sanskrit Buddhist terms and mantras, as well as the ways Chinese words were written in foreign alphabets. The writing system of

22  Introduction Chinese does not reflect the phonetic value, but transliteration data gives us a hint as to the phonetic value of ancient Chinese. 1.6.2.2.2  SINO-JAPANESE AND SINO-VIETNAMESE

Sino-Japanese refers to the reading of Chinese characters in Japan. Kan-on 漢 音 is believed to be a reflection of the Chang’an dialect in the eighth century, when the Japanese government sent ambassadors, students and monks to Tang China. An older tradition of Sino-Japanese is known as Go-on 吳音, which is probably based on the classical pronunciation of Chinese character in the Wu 吳 region during the Southern Dynasties. Sino-Vietnamese refers to the reading of Chinese characters in Vietnam. According to Maspero (1920), Sino-Vietnamese was primarily based on the Chang’an dialect at the end of the Tang dynasty. SinoJapanese and Sino-Vietnamese can provide a valuable reference point in the study of the origin of Sino-Korean, though it should also be noted that Sino-Japanese and Sino-Vietnamese are also loanword systems, which are undoubtedly subject to adjustment in the processes of adaptation.

1.7  Research structure and organization The structure of this book is organized as follows. Chapter 1 discusses the purpose, methodology and data of this study, as well as background information necessary to research. Chapter 2 introduces the theory of loanword phonology and investigates the mechanisms by which loanwords are adapted through a case study of Mandarin loanwords in Korean. Chapter 3, Chapter 4 and Chapter 5 investigate the initials, finals and tones of Middle Sino-Korean, respectively. Chapter 6 compares the sound system of Middle Sino-Korean and Middle Chinese as shown in different materials, paying special attention to those features which could aid our understanding of the origin of Sino-Korean. Chapter 7 is the conclusionary section of this book. We will summarize the conclusions reached in this book and outline tasks for future research.

Notes 1 The book Etudes sur la phonologie chinoise was published in 1926. The version that we are referring to is the Chinese version of this book, which was translated by Chao Yuenren et al. in 1940. 2 “The use of characters started since the beginning of the kingdom, and there was one who recorded affairs in one hundred volumes, which was named Yuki [Extant records]. 國初始用文字, 時有人記事一百卷, 名曰留記。 ” 3 “In the second year of king Sosurim, in the sixth month in summer, the king Fu Jian dispatched envoy and monk Shun Dao to send a figure of Buddha and Buddhist sutras (to Koguryo). 二年夏六月, 秦王符堅遣使及浮屠順道送佛像經文。 ” 4 “There was no written history record in Baekje since the beginning of the kingdom. By this time the savant of the Grand Academy Go Heung compiled the history book Seoki [written record]. 百濟開國已來, 未有以文字記事, 至是得博士高興, 始有書記。”

Introduction 23 5 “In the 27th year of king Geunchogo, in the first month in spring, (Baekje) sent envoy to present tribute to the Jin Dynasty. 二十七年春正月, 遣使入晉朝貢。” 6 “In the second year of Putong reign, the King, whose family name is Mo and given name Jin, started to send people along with Baekje’s envoy to present tribute. 普通二 年, 王姓慕名秦, 始使人隨百濟獻方物。 ” 7 “In the fourth year of the Wude reign in Tang, Silla sent envoy to present tribute. Since then Silla never stopped to present tribute to Tang. 唐武德四年, 新羅遣使朝貢 . . . 自 此朝貢不絕。” 8 “In the ninth year of King Seondeok, in the fifth month in summer, the king sent juniors to Tang, requesting to allow them attending the imperial college. 九年夏五月, 王遣子 弟于唐, 請入國學。” 9 “After Silla served Tang, Silla always sent princes to join Tang’s imperial guard troops and sent students to study in Tang’s imperial college. When these students returned Silla after the ten-year limit, Silla sent other students to Tang and the number of students was as many as more than one hundred. 新羅自事唐以後, 常遣王子宿衛, 又遣 學生入太學習業, 十年限滿還國, 又遣他學生入學者, 多至百餘人。 ” 10 This is a war between Korea and Japan in the year of 1592, which is an imjin 壬辰 year in the sexagenary cycle. 11 凡本國語音有平有仄 . . . 哀而安者為平聲, 厲而舉者為上聲, 直而高者為去聲入 聲。 12 The principle of fanqie is to notate the pronunciation of one character with two characters, with the first one indicating the initial, with the second indicating the rhyme. For example, the fanqie spelling for dong 東 is dehong 德紅. This shows that dong 東 has the same initial as de 德, and the same rhyme as hong 紅.

2 Mandarin loanwords in Modern Korean

In order to propose a convincing hypothesis on the adaptation strategy of SK, in this chapter, we will conduct a case study on the Mandarin loanwords in Modern Korean. Since the late twentieth century, Modern Korean has borrowed many Chinese proper names, such as brand names and place names, based on the pronunciation in Mandarin. Below are some examples. Table 2.1  Examples of Mandarin loanwords in Korean Word

Meaning

Mandarin

Adapted Form

Sino-Korean

烏龍茶 哈爾濱 元

oolong tea Harbin Chinese currency

u.lung.tʂʰa xa.ər.pin yɛn

u.lung.tɕʰa ha.ʌl.pin y.an

o.jong.tɕʰa ha.i.pin wʌn

In this chapter, we will analyze the adaptation of Mandarin loanwords from the perspective of loanword phonology. We are optimistic that the result of this study will aid us to better understand the mechanisms by which loanwords are adapted as well as many aspects of the nature of Sino-Korean phonology. Before we proceed to our discussion of the adaptation of Mandarin loanwords, in the following sections, we will briefly introduce Mandarin and Korean phonology.

2.1  Mandarin phonology The Mandarin phonological system will be discussed in the following order: consonants, vowels, syllable structure and tones. 2.1.1  Mandarin consonants 2.1.1.1  The number of consonants Most scholars agree that modern Mandarin has 22 consonants (e.g., Cheng 1973; Dong 1996), whereas Lin (2007) and Duanmu (2000) claim that the -j [ʨ], -q [ʨʰ] and -x [ɕ] are not consonant phonemes in Mandarin. Lin (2007: 48) notes that

Mandarin 25 “these three consonants belong to one set of consonants that is in complementary distribution with the three other sets of consonants: the dental affricates/fricatives, the post-alveolar affricates/fricatives, and the velars,” and “these three consonants are not independent phonemes but allophones that occur only before high front vowels/glides.” In this book, we will not follow Lin’s (2007) viewpoint for the following two reasons. First, as Cheng (1973) states, -j [ʨ], -q [ʨʰ] and -x [ɕ] can be treated as distinctive phonemes in modern Mandarin given that there is no compelling phonological evidence in choosing either -z [ts], -c [tsh] and -s [s] or -g [k], -k [kh] and -h [x] as the phoneme of -j [ʨ], -q [ʨʰ] and -x [ɕ]. Second, one of the principles to determine whether a sound is a phoneme is to judge if the sound is perceived to be different by native speakers. For example, there is no minimal pair for the sounds [h] and [ŋ] in English, and these two sounds can be shown to be in complementary distribution. Although this fact can be used to argue that they are allophones of the same phoneme, they are so different phonetically that they are still considered different phonemes. For native Mandarin speakers, the sounds -j [ʨ], -q [ʨʰ] and -x [ɕ] are reasonably different from -z [ts], -c [tsh] and -s [s] as well as -g [k], -k [kh] and -h [x]; therefore, it is reasonable to treat them as distinctive phonemes. 2.1.1.2  The sound value of Mandarin consonants One major debate concerning the sound value of Mandarin consonants relates to the sound which is presented by -r in Hanyu pinyin, which is by some viewed as a voiced fricative [ʐ] (Karlgren 1926[1940]; Dong 1996) and by others an approximant [ɹ] (Fu 1956; Wang 1983; Duanmu 2000; Lin 2007) or a retroflex approximant [ɻ] (Zhu 1982). In addition, some scholars claim that the phonetic value of the Mandarin -r is pronounced differently in different phonetic environments. The sound has two allophones in terms of frication, namely, voiced fricative [ʐ] and frictionless continuant [ɻ]. However, in terms of manner of articulation, the Mandarin -r can be described as [ɹ] (Liao and Shi 1987). Treating the r as [ʐ] has an advantage in that it can be paired with [ʂ], which is distinguished by [+/- voiced]. However, it also causes a problem in the phonemic system of Mandarin, since there are no other voiced fricatives. In this study, we will transcribe the pinyin -r symbol as [ɻ] in IPA. 2.1.1.3  The place of articulation of consonants First, Lin (2007: 41) claims that the Mandarin [t], [tʰ], [n] and [l] can be dentals or alveolar or either one, depending on the speaker. In Lin (2007), they are treated as dentals, just like -z [ts], -c [tsʰ] and -s [s]. However, Dong (1996) claims that generally [t], [tʰ], [n] and [l] are more similar to alveolar, while -z [ts], -c [tsʰ] and -s [s] are more similar to dentals. In this study, we will follow Dong (1996) to treat [t], [tʰ], [n] and [l] as alveolars and [s], [ts] and [tsʰ] as dentals. Second, zh [tʂ], ch [tʂh] and sh[ʂ] have been referred to as retroflex sounds since Chao (1948, 1968). However, Lin (2007) states that phonetic studies have

26  Mandarin shown that these sounds are not retroflexes in the normal sense in the speech of the speakers of the Beijing dialect. Lin (2007) named the -zh [tʂ], -ch [tʂh] and -sh [ʂ] as post-alveolar consonants. Some scholars claimed that Standard Chinese speakers produce the constriction for -zh [tʂ], -ch [tʂh] and -sh [ʂ] with the upper surface of the tip of the tongue. This makes these sounds laminals rather than apical post-alveolars (Ladefoged and Maddieson 1996: 150–154). In this study, we will follow the traditional viewpoint to treat the -zh [tʂ], -ch [tʂh] and -sh [ʂ] as retroflexes. 2.1.1.4 Summary Based on the above discussion, the inventory of Mandarin consonants can be listed as below. Table 2.2  Inventory of Mandarin consonants Bilabial Labiodentals DentoAlveolar Retroflex Palatal Velar Alveolar Plosive p pʰ Nasal m Fricative Affricate Approximant Lateral Approximant i

f

s ts tsʰ

t tʰ n

l

ʂ tʂ tʂʰ ɻ

ɕ tɕ tɕʰ

k kʰ ŋi x

[ŋ] occurs only in the coda position.

2.1.2  Mandarin vowels Mandarin vowels can be divided into monophthongs and diphthongs. A monophthong is a vowel consisting of one articulation from the beginning to the end, whereas a diphthong is a vowel including two articulations. 2.1.2.1  Mandarin monophthongs According to a traditional analysis, at the surface representation (SR), there are 12 Mandarin vowels, which are [a, ɑ, ɛ, e, ɤ, o, ə, ʅ, ɿ, i, y, u] (Cheng 1973). From a practical perspective, as reflected in the pinyin system, Mandarin has six vowels in its underlying representation (UR), namely, [a, o, e, i, u, y]. However, there exists much disagreement concerning the phonemic nature of Mandarin monophthongs, and we will discuss this in the following section. 2.1.2.1.1  HIGH VOWELS

The two high vowels [i] and [u] are accepted as phonemes without debate. However, scholars have long disagreed as to whether the Mandarin vowel ü [y] is a

Mandarin 27 monophthong or diphthong. Scholars such as Chao (1968) claim that [y] should be considered a complex segment, which is composed of [ui] or [iu]. However, scholars such as Cheng (1973) claim that treating [y] as a complex segment not only has no historical evidence, but also is not supported by linguistic proofs. In this book, we take the vowel [y] as a monophthong from the perspective that when articulating this vowel, there is no particular tongue movement that would suggest a diphthongal pronunciation. Another much-debated question is the sound value of the two apical vowels [ɿ] and [ʅ]. Many scholars treat these two vowels as apical vowels (Karlgren 1926; Cheng 1973; Dong 1996). However, considering their characteristics, such as dental articulation and retroflexion, some scholars regard these two vowels as syllabic consonants (e.g., Fu 1956; Pulleyblank 1983). Duanmu (2000) treats them as syllabic consonants ([z], [r]). Lin (2007) argues that the two apical vowels are the syllabic consonant [ɹ̩ ] and function like vowels. In this study, we will treat them as vowels for the convenience of description: If the [ɿ] and [ʅ] are not vowels, it would be difficult to explain the correspondence between these two vowels and the Korean vowel [ɯ] in Mandarin loanwords in Korean. Since using one symbol [ɹ̩ ] to represent two vowels may cause confusion, we will use the IPA symbols [ɿ] and [ʅ] to distinguish these two vowels clearly. Scholars have different opinions about the underlying form of the two apical vowels [ɿ] and [ʅ]. The vowel [ɿ] appears only after the dental affricates and fricatives in syllable final position, while the [ʅ] appears only after retroflex (postalveolar) consonants in syllable final position. The places of articulation of these two vowels are the same as that of the preceding consonant. Cheng (1973) argues that the two vowels are the allophones of the mid high vowel [ɨ]. However, some scholars view the [ɿ], [ʅ] and [i] as the allophones of the underlying [i], since these three sounds are in complementary distribution. Given the predictability of the occurrence of [ɿ] and [ʅ], in this book, we will regard [i] as the UR of [ɿ], [ʅ] and [i]. 2.1.2.1.2  MID VOWELS

The surface forms of Mandarin mid vowels include [e, o, ɤ, ə]. Lin (2007: 76) states that there is only one mid vowel in Mandarin, namely [ə], based on the distribution of these four vowels: [ə] before consonants [e] in a diphthong followed by [i] or in syllable final position preceded by [j]/ [ɥ] [o] in a diphthong followed by [u] or in syllable final position preceded by [w] [ɤ] in syllable final position in a CV syllable As shown above, all of these four mid vowels appear in different phonetic environments. The appearance of [e, o, ɤ] is predictable if the underlying form is set as [ə]. Therefore, in this book, we will follow Lin (2007) to use [ə] as the underlying phoneme for the Mandarin mid vowels.

28  Mandarin 2.1.2.1.3  LOW VOWELS

In the surface form, there are two low vowels in Mandarin, [a] and [ɑ]. However, these two low vowels are in complementary distribution, as shown below. [a] in the syllable final position, diphthongs ending in [i] or before [n] [ɑ] in diphthongs ending in [u] or before [ŋ] Since [ɑ] occurs in predicable contexts, we will treat [a] as the unique low vowel in the underlying form. In addition, when the low vowel is between [j]/[ɥ] and [n], the vowel becomes fronted and raised to [ɛ]. The [ɛ] sound occurs in a predictable phonetic environment, so we will choose [a] as the underlying form of the sound [ɛ]. 2.1.2.1.4  RETROFLEX VOWEL

The retroflex vowel [ɚ] occurs in the rime er in Mandarin. It is also called rhotacized (r-colored) vowels or rhotic vowels. Scholars have long debated whether this retroflex vowel [ɚ] is an independent phoneme. It is regarded as an independent vowel in Dong (1996: 16) but not in Duanmu (2000) and Lin (2007). In this book, we agree that this retroflex vowel [ɚ] is not a phoneme for the following reasons: First, no onset consonant can appear before the rime er. Second, the retroflex or the rhotacization vowel is not an independent vowel, rather merely a vowel possessing a rhotacized feature, which is a feature that can be added to many different vowels in Mandarin. For example, the morpheme ba “handle” becomes bar “handle” after the r-suffixation. 2.1.2.1.5 SUMMARY

Based on the discussion above, the inventory of Mandarin vowels in both SR and UR can be listed as below. Table 2.3  Inventory of Mandarin vowels (SR) Front

High Mid Mid-Low Low

Central

unrounded

rounded

i, ɿ e ɛ

y

Back unrounded

rounded

ə

ʅ ɤ

u o

a

ɑ

Central

Back

Table 2.4  Inventory of Mandarin vowels (UR) Front

High Mid Mid-Low Low

unrounded

rounded

i

y

unrounded ə a

rounded u

Mandarin 29 2.1.3  Mandarin syllables 2.1.3.1  Structure of Mandarin syllables In traditional Chinese phonology, tone is considered a property of the whole syllable that affects the whole of the Chinese syllable, both initial and final. An initial is a syllable-initial consonant, and the place of the initial in Mandarin can be unoccupied, which is referred to as a zero-initial. A  final is composed of a medial and a rhyme. The medial in Chinese is the glide before the nuclear vowel, including [i], [u] and [y], or, more accurately, [j], [w] and [ɥ]. The rhyme contains a nucleus and an ending. The nucleus is the main vowel, and it is the most essential element in a Mandarin syllable. In other words, a nucleus must exist in a Mandarin syllable, while all the other elements are optional. The ending can be either the vowels [i/u] or the nasals [n/ŋ]. The structure of a Mandarin syllable can be drawn as below. Table 2.5  Traditional description of Mandarin syllable structure Initial

Final Medial

(C)

(G) i, u, y

Rhyme Nucleus V TBU (Tone Bearing Unit)

Ending (V/C) i, u, n, ŋ

Although the traditional description of Mandarin syllable structure places the medial under the unit of the final, some scholars (e.g., Duanmu 1990, 2000; Lin 2007) claim that a medial should be placed in the onset position of a syllable. Duanmu (1990) provides the evidence that the Mandarin [hu] and [f] are changed to [f] and [hu] in the Santai dialect in Sichuan. For example, the Mandarin [huei] “ash” is pronounced as [fei] and Mandarin [fei] “fly” as [huei] in Santai. In this book, we will follow the traditional viewpoint that states that a medial should be placed under a final for the convenience of comparing the syllable structures between Korean and Mandarin. Based on the discussion above, the structure of a Mandarin syllable is (C)(G)(V)(C/V). The longest syllable is CGV(C/V), and the shortest syllable is V. Table 2.6 lists all the possible types of mandarin syllables. The examples are cited from Lin (2007). Table 2.6  Syllable types in Mandarin

1. V1 2. CV 3. GV 4. CGV 5. VC

Pinyin

IPA

Character

Meaning

è mǎ yā duō ān

[ɤ]51 [ma]214 [ja]55 [two]55 [an]55

餓 馬 鴨 多 安

hungry horse duck many peace, safe (Continued)

30  Mandarin Table 2.6 (Continued)

 6.VV   7. CVC   8. CVV   9. GVC 10. GVV 11. CGVC 12. CGVV

Pinyin

IPA

Character

Meaning

ài lán lái yán yào huàn huài

[ai]51 [lan]35 [lai]35 [jɛn]35 [jɑu]51 [xwan]51 [xwai]51

愛 藍 來 鹽 藥 換 壞

love blue come salt medicine to exchange bad

2.1.3.2  Phonotactic constraints in Mandarin There are phonotactic constraints on what sequences of segments can be combined in Mandarin syllables. Since not all initials can combine with every final, the real number of Mandarin syllable is less than the number of potential syllables in Mandarin. Some constraints on the combination of initial and finals are shown as below. a b c d

Bilabials [p, pʰ, m], alveolars [t, tʰ], retroflexes [tʂ, tʂʰ, ʂ] and dento-alveolars [ts, tsʰ, s] cannot combine with the high front round vowel [y]. The labiodentals [f] and velars [k, kʰ, x] cannot combine with [i, y]. Palatals [ʨ, ʨʰ, ɕ] can only combine with [i, y]. The [ŋ] consonant cannot appear in the onset position.

In addition to the syllables prohibited by such restrictions, some particular types of Mandarin syllables are not seen in Mandarin. This can be viewed as an accidental gap. The following syllables do not exist in Mandarin, for example: [ɥa], [ɥɑŋ], [ɥəŋ], [kwəŋ], [tia], [nia], [tin], [tiɑŋ], [lwei] and [nwɑŋ] (Lin 2007: 120). 2.1.4  Mandarin tones In Mandarin, there are four lexical tones, namely 55, 35, 214 and 51, according to the five-point scale devised by Chao (1930), in which 1 indicates the lowest pitch and 5 indicates the highest pitch. The most well-known examples of the four tones in Mandarin are shown in Table 2.7. Table 2.7  Four tones in Mandarin Tone name

Phonological description

Chao tone numerals

Chinese character

Gloss

Tone 1 Tone 2 Tone 3 Tone 4

High level Mid rising Low dipping High falling

ma 55 ma 35 ma 214 ma 51

媽 麻 馬 罵

mother hemp horse scold

Mandarin 31 In addition to these four phonemic tones, there is one neutral tone. The neutraltone syllable is an unstressed syllable shorter than a normal syllable. For example, the suffix zi in yizi “chair” is a neutral tone syllable, which is shorter than the zi in the word zinü “son and daughter.” The pitch of the neutral tone is characterized as half-low after /H/ [55], middle after /LH/ [35], half-high after /L/ [21] and low after /HL/ [51] (Chao 1968).

2.2  Korean phonology 2.2.1  Korean consonants There are six major dialects in Korean, which are the Central dialect, Southeast dialect, Southwest dialect, Northeast dialect, Northwest dialect and Jeju dialect (Pak 2009). In this book, we will introduce the phonological system of Standard Korean, known as “Standard Korean Pronunciation (SKP).” Although the SKP is highly similar to the phonological system of the Seoul Korean, they are not completely identical. “Standard Korean Pronunciation (SKP)” is based on the document “Pyojun bareum beop” 표준발음법 [Standard pronunciation law] published in 1988 and implemented in 1989 by the Korean government. Many scholars have indicated that SKP is marginally different from the actual speech of Seoul speakers (e.g., Lee and Ramsey 2000; Shin et al. 2012). In the following section, we will introduce SKP first and discuss the differences between SKP and Seoul Korean where necessary. According to documentation on SKP, there are 19 consonants in Korean. This is also true in Seoul Korean (e.g., Shin et al. 2012; Sohn 1999; Lee and Ramsey 2000). However, there is much disagreement about how to transcribe some of the 19 consonants. The first controversy concerns the tense consonants2 -ppㅃ, -tt ㄸ and -kkㄲ in Korean. In the IPA Handbook (1999) these consonants are transcribed as [b, d, g]. However, under the heading of conventions, it says that these [b, d, g] are voiceless unaspirated (or slightly aspirated) lenis plosives [b̥ , d̥ , g̊ ] in the beginning of a syllable, but are regularly realized as voiced sounds in intervocalic position. Shin et al. (2012) shows that in phonetic graphs, none of these consonants are accompanied with any vibration syllable-initially and these three tense consonants become [b/b̥ , d/d̥ , g/g̊ ] in intervocalic position. In other words, the voicing of these three tense consonants in the intervocalic position is the result of assimilation, which is predictable, and the underlying form is still voiceless. Therefore, the [b, d, g] symbols, which notate voiced consonants conventionally, will not be used to transcribe the tense consonants in Korean in this research. In this study, we will follow the notation in Shin et al. (2012), which transcribes Korean tense consonants as [p*, t*, k*]. Second, the transcription of the palatal consonants -cㅈ, -ccㅉ, -chㅊis controversial among scholars. Shin et al. (2012) transcribe these consonants as [tɕ, tɕ*, tɕʰ], while the IPA handbook (1999) transcribes these consonants as [ɟ, c, cʰ]. In fact, however, the transcription in the IPA handbook is inconsistent. Specifically, on page 120 of IPA handbook (1999) the [c, cʰ, ɟ] in Korean is placed under the

32  Mandarin “affricate” column, but these symbols [c, cʰ, ɟ], according to the IPA chart in page ix, are listed as plosive consonants. Given this inconsistency, we choose not to adopt the transcription of the IPA handbook (1999), but rather will follow Shin et al.’s (2012) transcription of the palatal consonants in Korean. Based on the discussion above, the inventory chart of Korean consonants can be listed as below. Table 2.8  Inventory of Korean consonants

Plosive Fricative Affricate Nasal Liquid

Lax Tense Aspirated Lax Tense Lax Tense Aspirated

Bilabial

Alveolar

p (ㅂ) p* (ㅃ) pʰ (ㅍ)

t (ㄷ) t*(ㄸ) tʰ (ㅌ) s (ㅅ) s*(ㅆ)

m (ㅁ)

n (ㄴ) l (ㄹ)

Alveolopalatal

Velar k (ㄱ) k* (ㄲ) kʰ (ㅋ)

tɕ (ㅈ) tɕ* (ㅉ) tɕʰ (ㅊ)

Glottal

h (ㅎ)

ŋ (ㅇ)

2.2.2  Korean vowels In terms of vowels, the SKP is slightly different from the actual speech of Seoul speakers. The vowel system of SKP and that of the Seoul Korean will be introduced in the following sections immediately below. 2.2.2.1 Monophthongs According to documentation on SKP, there are ten monophthongs in Standard Korean, which are listed as follows: Table 2.9  Monophthong system of Standard Korean Front

High Mid Low

Back

unrounded

rounded

unrounded

rounded

iㅣ eㅔ æㅐ

yㅟ øㅚ

ɯㅡ ʌㅓ ɑㅏ

uㅜ oㅗ

However, this vowel system is not identical to actual speech in the Seoul dialect of Korean in the following aspects: First, according to SKP, the two vowels -wi3 ㅟ [y] and -oyㅚ[ø] are monophthongs, although they can be pronounced as diphthongs. For example, the word /ø-kuk/ 외국 “foreign country” can be pronounced as either [ø-kuk] or [ue-kuk]. The word /y/ 위 “up, above” can be pronounced

Mandarin 33 as either [y] or [wi]. However, in the Seoul Korean, -wi ㅟ and -oy ㅚ are not monophthongs. According to Shin et al. (2012) native speakers of Seoul Korean pronounce -wi ㅟ [y] and -oy ㅚ [ø] as diphthongs. In the spectrogram for these two vowels, there is formant transition in both figures. This is also pointed out in the IPA handbook (1999: 121). Second, according to SKP the two vowels [e] ㅔand [æ] ㅐ are distinguishable. Orthographically, the word [ke] 게 “crab” and [kæ] 개 “dog” are different, and it is expected that these two words should sound different. However, scholars have found, to the contrary, many native speakers cannot distinguish these two sounds. Specifically, Choe (2002) shows that among 210 native speakers of Seoul Korean between 20 and 80 years old, more than 80 percent of the time participants pronounced the two sounds identically. Shin et al. (2012) also shows no significant difference can be observed between [e]ㅔ and [æ] ㅐ in their formant-measures F1 and F2 in their experiment, which used information gathered from ten male and ten female native speaker participants. Shin et al. (2012) concludes that there are only seven monophthongs in Seoul Korean speech, namely [i], [ɯ], [u], [ɛ], [ʌ], [o] and [ɑ]. The debate on the phonetic value of Korean monophthongs mainly concerns the following sounds. First, the transcription for mid vowel -eyㅔand -ayㅐare [e] and [ɛ] in some studies (e.g., IPA handbook 1999; Lee and Ramsey 2000; Lee and Kim 2010), but mid vowel [e] and low vowel [æ] in Shin et al. (2012). Shin et al. (2012) argues that these two are not distinguishable, and they use the mid-low vowel [ɛ] to represent both of them in the phonemic inventory. Lee and Ramsey (2000: 64) state that “young speakers in Seoul are seldom able to actually hear the difference. Because the two vowels are spelled differently in Hangul, most literate Koreans know which value a word is supposed to have from its spelling. But how they actually speak is something else.” Moreover, “speakers of standard Korean often cannot distinguish the English words pet and pat, men and man, bend and band.” In this study, we will follow the transcription provided in Shin et al. (2012). In the orthography of Korean, the -eyㅔand -ayㅐare represented differently, as shown in the Yale Romanization, in which they are written as -ey and -ay, respectively. Since it is generally an accepted trend that orthography reflects the pronunciation of an earlier period, we will respect this difference and view it as the feature of the underlying form of these two vowels. In IPA, both [e] and [ɛ] are mid vowels. So in order to distinguish these two vowels in our vowel chart of the Prescriptive Korean, namely SKP, we will transcribe them as mid vowel [e] and low vowel [æ] respectively. However, in the monophthong system of the Seoul Korean both -eyㅔand -ayㅐwill be transcribed as [ɛ]. Second, the transcription for eㅓ is [ə] in Lee and Ramsey (2000), but [ʌ] in the IPA handbook (1999) and Shin et al. (2012). In Lee and Kim (2010), the transcription for -eㅓis [ə] in the word eon-je 언제 “when,” but [ʌ] in the word meo-ri 머리 “head.” Since the occurrence of [ə] in the word eon-je 언제 “when” is in predicable context (being followed by the coda [n]), in this study we will choose [ʌ] as the transcription for mid vowel -eㅓ in Korean. Third, the high unrounded vowel -u ㅡ is transcribed as [ɨ] in Lee and Ramsey (2000), but [ɯ] in the IPA handbook (1999), Lee and Kim (2010) and Shin et al.

34  Mandarin (2012). Although Lee and Ramsey (2000) transcribed -u ㅡ as [ɨ], they claim that this sound is the unrounded equivalent of [u] (Lee and Ramsey 2000: 65). In the IPA chart, the unrounded equivalent of [u] should be [ɯ]. Following this, we will use the symbol [ɯ] to transcribe the high unrounded vowel -u ㅡ in Korean. Fourth, the -aㅏ is transcribed as the central [a] in many studies (IPA handbook 1999; Lee and Ramsey 2000; Lee and Kim 2010). But it is transcribed as the back [ɑ] in Shin et al. (2012). Since there is only one /a/ in Korean, and the back [ɑ] occurs in a predictable phonetic environment, we will choose the central [a] as the phoneme. Based on our discussion above, the monophthong system of the Seoul Korean can be listed as follows: Table 2.10  Monophthong system of the Seoul Korean Front

Central

unrounded High Mid Low

Back

rounded

i ɛ

a

unrounded

rounded

ɯ ʌ

u o

2.2.2.2 Diphthongs Diphthongs can be divided into on-glide diphthongs and off-glide diphthongs based on the order in which the glide appears. An on-glide diphthong is one where the glide precedes the monophthong, and an off-glide is one where the glide follows the monophthong. According to SKP, there are eleven diphthongs, which are the combination of the ten monophthongs and two glides [j] and [w]. There is no off-glide diphthong in Modern Korean. Table 2.11  Diphthong system of the Seoul Korean

j w

i

ɛ

æ

a

ɯ

u

ʌ

o

-

je ㅖ we ㅞ

jæ ㅒ wæ ㅙ

jaㅑ wa ㅘ

ɯi ㅢ -

-ju ㅠ -

jʌ ㅕ wʌ ㅝ

jo ㅛ -

As we mentioned above, SKP is an idealized or prescriptive system of Korean pronunciation, which is different from actual speech in Seoul Korean. The -eyㅔ and -ayㅐare undistinguishable in Seoul Korean; consequently, the -je ㅖis undistinguishable from -jæ ㅒ, and -we ㅞ undistinguishable from -wæ ㅙ. Moreover, the -wi ㅟ and -oy ㅚ are considered as monophthongs in SKP, but are diphthongs in Seoul Korean. 2.2.3  Korean syllables Traditionally, a Korean syllable is divided into three parts: choseong 初聲, which are syllable-initial consonants, jungseong 中聲, which corresponds to the medial and nucleus, and jongseong 終聲, which corresponds to the coda position.

Mandarin 35 Table 2.12  Syllable structure of Korean Choseong

Jungseong

Initial (C)

Medial (G)

Jongseong Nucleus V

Coda (C)

The syllable structure in Korean can be summarized as (C)(G)V(C), in which only the nuclear vowel is required. Consonant clusters are not allowed in Korean. The choseong (initial) part can be a zero-initial or any one of the nineteen consonants.4 The jungseong part can be one of the ten vowels or eleven diphthongs. The coda position can be occupied by one of the seven consonants: -p, -t, -k, -m, -n, -ŋ and -l. The following chart shows all the syllable types in Korean. Table 2.13  Syllable types in Korean

1.V 2.CV 3.GV 4.CGV 5.VC 6.CVC 7.GVC 8.CGVC

IPA

Hangul

Meaning

i na wa hjo an sam yak hwal

이 나 와 효 안 삼 약 활

teeth I and filial duty inside three medicine bow

2.2.4  Korean vowel length and pitch There is no tone in Standard Korean. In the SKP the vowel length in some words can distinguish meaning. The macron diacritic in the following examples indicates the vowel is long: mal “horse” vs. ma̅ l “speech”; nun “eye” vs. nu̅ n “snow”; pam “night” vs. pa̅ m “chestnut.” Vowel length is maintained only in the first syllable of a word. If a word with a long vowel appears in non-initial position in a compound, the vowel becomes a short vowel (Lee and Ramsey 2000). In some dialects, the difference between the first and the second type is not reflected in vowel length, but in pitch. The first type has a higher pitch, and the second type a lower pitch (Pak 2009).

2.3  Korean transcription systems for Mandarin loanwords 2.3.1  South Korean government transcription In 1986 the South Korean Ministry of Education announced the official document Oeraeeo pyogibeop 외래어 표기법 “Rules for writing foreign language words in Hangul” to dictate how foreign words should be adapted and transcribed into Korean.5 The basic rules proposed in this document are as listed below. a b

Only the modern set of Hangul alphabet should be used. One phoneme should be represented by only one symbol.

36  Mandarin c

For the coda sounds, only the following alphabetic symbols are used in the orthography: ㄱ[k], ㄴ[n], ㄹ[l], ㅁ[m], ㅂ[p], ㅅ[t],6ㅇ[ŋ]. d A plosive should not be represented with a tense consonant (e.g., ㄲ[k*], ㄸ [t*], ㅃ[p*], ㅆ[s*], ㅉ[ts*]). e Those loanword transcriptions that are already widely in use should be respected. In addition to these general rules, there are two particular rules with regard to the transcription of Mandarin loanwords. a b

Mandarin tones will not be transcribed. Mandarin [ia], [ie], [io] and [iu] are written as ㅏ[a], ㅔ[e], ㅗ[o] and ㅜ[u] when the preceding Mandarin initials are adapted in Korean as ㅈ[tɕ], ㅉ[tɕ*] or ㅊ[tɕʰ].

In addition to the Korean government transcription rules, there are some other proposals on the transcription of Mandarin loanwords in Korean. Below, we will introduce the Korean government transcription first and a discussion on other proposals will follow. 2.3.1.1 Consonants 2.3.1.1.1 PLOSIVES

The following table lists the plosive consonants in Mandarin and Korean. Table 2.14  Plosive consonants in Mandarin and Korean Bilabial voiceless unaspirated Mandarin p Lax Tense Korean p p*

Alveolar voiceless voiceless aspirated unaspirated pʰ t Lax Tense pʰ t t*

Velar voiceless voiceless aspirated unaspirated tʰ k Lax Tense tʰ k k*

voiceless aspirated kʰ kʰ

Mandarin has two pairs of plosives, namely, voiceless unaspirated and voiceless aspirated plosives, while Korean has a three-way contrast in its plosive system, namely, lax unaspirated, tense unaspirated, and aspirated plosives. The Mandarin voiceless aspirated [pʰ], [tʰ] and [kʰ] are perfectly mapped into the voiceless aspirated [pʰ] ㅍ, [tʰ] ㅌ and [kʰ] ㅋ in Korean, respectively, as shown below. Table 2.15  Korean transcription of Mandarin aspirated plosives Pinyin

IPAi

Character

Hangul

IPA

pǔěrchá máotáijiǔ kǒngfǔjiājiǔ

pʰu. ər. tsʰa mɑu.tʰai. tɕjəu kʰuŋ. fu. tɕia. tɕjəu

普洱茶 茅臺酒 孔府家酒

푸얼차 마오타이주 쿵푸자주

pʰu.ʌl.tɕʰa ma.o. tʰa. i. tɕu kʰuŋ.pʰu. tɕa. tɕu

i  For simplicity, tone notation is not included in the transcription.

Mandarin 37 However, the transcription of Mandarin voiceless unaspirated -b [p], -d [t], -g [k] poses a challenge to scholars. In Chinese there is only one series of voiceless unaspirated plosives, while in Korean there are two series, namely lax voiceless plosives ㅂ[p], ㄷ[t], ㄱ[k] and tense voiceless plosives ㅃ[p*], ㄸ[t*], ㄲ[k*]. Phonetic differences between Korean tense and lax stops have been discussed in many studies. Below is a summary of the phonetic differences between tense and lax stops in the initial position given in Kim and Duanmu (2004). Table 2.16  Phonetic differences between Korean tense and lax stops Feature

Tense

Lax

Following tone VOT Glottal opening H1-H2 (breathiness) Intensity Voicing duration Airflow at release Air pressure before release

higher shorter narrower smaller strong shorter smaller greater

lower longer wider larger weak longer greater smaller

Mandarin -b [p], -d [t], -g [k] and Korean ㅂ[p], ㄷ[t], ㄱ[k] are represented by the same IPA symbols; however, many studies show that the Mandarin -b [p], -d [t], -g [k] are perceived to be more similar to the Korean tense stops [p*], [t*] and [k*] than the lax stops (Eom 1996; Jeon 1999). Despite the perception similarity with Korean tense stops, the Mandarin -b [p], -d [t], -g [k] are adapted as Korean lax stops ㅂ[p], ㄷ[t], ㄱ[k] in the South Korean government transcription. Table 2.17  Korean transcription of Mandarin unaspirated plosives Pinyin

IPA

Character

Hangul

IPA

běijīng dàlián gānsù

pei.tɕiŋ ta.liɛn kan.su

北京 大連 甘肅

베이징 다롄 간쑤

pe.i. tɕiŋ ta.lien kan.s*u

2.3.1.1.2 FRICATIVES

The following table compares the fricatives in Mandarin and Korean. Table 2.18  Fricative consonants in Mandarin and Korean Labio-dental Dental Alveolar Mandarin f Korean

Retroflex Alveo-palatal Velar Glottal

s ʂ lax tense s s*

ɕ

x h

The labio-dental [f] sound in Mandarin does not have any corresponding sound in Korean. The Korean bilabial aspirated [pʰ] is used to transcribe the Mandarin [f] in the government transcription.

38  Mandarin Table 2.19  Korean transcription of Mandarin [f] Pinyin

IPA

Character

Hangul

IPA

fújiàn

fu.tɕjɛn

福建

푸젠

pʰu. tɕen

Similar adaptation is also seen in the English loanwords in Korean. The English voiceless fricative [f] always corresponds to the [pʰ] in Korean, as exemplified below. Table 2.20  Korean transcription of English [f] English

Hangul

IPA

film foundation coffee muffler

필름 파운데이션 커피 머플러

[pʰil.lɯm] [pʰa. un. tɛ. i. sjʌn]i [kʰʌ.pʰi] [mʌ. pʰɯl. lʌ]

i  “Foundation” is also adapted as [hwa. un. tɛ. i. sjʌn] in Korean.

The Mandarin fricatives -s [s] and -sh [ʂ] are represented as ㅆ[s*] and ㅅ[s], respectively, in Korean in the government transcription. Table 2.21  Korean transcription of Mandarin fricatives [s] and [ʂ] Pinyin

IPA

Character

Hangul

IPA

sìchuān shànghǎi

sɿ.tʂʰwan ʂɑŋ.xai

四川 上海

쓰촨 상하이

s*ɯ.tɕʰwan sɑŋ. ha. i

The Mandarin fricative -s [s] and Korean lax fricative -s ㅅ are both written as [s] in IPA. However, in terms of perception, the retroflex fricative -sh [ʂ] sounds more like the Korean lax fricative [s]ㅅ, while the Mandarin -s [s] sounds more similar to the Korean tense [s*]ㅆ (Eom 1996). The transcription of Mandarin fricatives -s [s] and -sh [ʂ] shows that perception plays a crucial role in the adaptation. The Mandarin alveo-palatal fricative -x [ɕ] corresponds to the lax fricative ㅅ[s] in Korean. Table 2.22  Korean transcription of Mandarin fricative [ɕ] Pinyin

IPA

Character

Hangul

IPA

xiàmén xīnjiāng

ɕja. mən ɕin. tɕjaŋ

廈門 新疆

샤먼 신장

ɕja. mʌn ɕin. tɕaŋ

According to Shin et al. (2012), the fricatives [s] and [s*] in Korean have the following allophones.

Mandarin 39 Table 2.23  Allophones of Korean fricative [s] and [s*] Phoneme

Major allophones

Environment

s

ɕ s ɕ* s*

before /j, i, wi/ before all other vowels before /j, i/ before all other vowels

s*

Since theㅅ[s] in Korean becomes [ɕ] when it is followed by [i] or [j], in this sense, the [ɕ] in Mandarin is faithfully mapped into Korean. The Mandarin velar fricative [x] is transcribed as the glottal fricative [h] ㅎ in Korean. Table 2.24  Korean transcription of Mandarin fricative [x] Pinyin

IPA

Character

Hangul

IPA

hǎiěr

xai.ər

海爾

하이얼

ha. i. ʌl

2.3.1.1.3 AFFRICATES

The following chart lists the affricates in Mandarin and Korean. Mandarin has three types of affricates, while Korean has only one type, namely palatal affricates. Table 2.25  Affricates in Mandarin and Korean

Mandarin Korean

Dental-alveolar

Retroflex

unaspirated ts

unaspirated tʂ

aspirated tsʰ

Palatal aspirated tʂʰ

unaspirated tɕ lax tense tɕ tɕ*

aspirated tɕʰ aspirated tɕʰ

All the three Mandarin aspirated affricates, regardless of their place of articulation, are represented as [tɕʰ], the only aspirated affricate in Korean. Table 2.26  Korean transcription of Mandarin aspirated affricates Pinyin

IPA

Character

Hangul

IPA

cāngzhōu chángshā qīngdǎo

tsʰɑŋ. tʂou tʂʰɑŋ. ʂa tɕʰiŋ. tɑu

滄州 長沙 青島

창저우 창사 칭다오

tɕʰaŋ. tɕʌ.u tɕʰaŋ. ɕa tɕʰiŋ. ta.o

The Mandarin palatal unaspirated affricate [tɕ] is represented as the lax affricate [tɕ] in Korean, which is also a palatal sound. Table 2.27  Korean transcription of Mandarin unaspirated affricate [tɕ] Pinyin

IPA

Character

Hangul

IPA

jǐnán

tɕi.nan

濟南

지난

tɕi. nan

40  Mandarin The Mandarin retroflex affricate -zh [tʂ] and dental-alveolar affricate -z [ts] are represented as lax affricate [tɕ] and tense affricate [tɕ*], respectively, in Korean. Table 2.28  Korean transcription of Mandarin unaspirated affricate [ts] and [tʂ] Pinyin

IPA

Character

Hangul

IPA

zībó zhāngjiājiè

tsɿ.po tʂaŋ. tɕja. tɕje

淄博 張家界

쯔보 장자제

tɕ*ɯ.po tɕaŋ.tɕa.tɕe

As we mentioned above, in terms of perception, the Mandarin retroflex fricative -sh [ʂ] is close to the Korean lax fricative [s] ㅅ, while the Mandarin -s [s] sounds more alike to the Korean tense fricative [s*] ㅆ. Likewise, the Mandarin -zh [tʂ] sounds more similar to the Korean lax affricate [tɕ], while the Mandarin -z [ts] sounds more like the Korean tense affricate [tɕ*]. The government transcription shows that perception plays a crucial role in the adaptation of the Mandarin affricative -zh [tʂ] and -z [ts]. 2.3.1.1.4 APPROXIMANTS

The following chart lists the approximants in Mandarin and Korean. Mandarin retroflex [ɻ] is not found in the approximant inventory of Korean. Table 2.29  Approximants in Mandarin and Korean Bilabial Mandarin Korean

Alveolar

Retroflex

l l

ɻ

Palatal

The two Mandarin approximants, -l [l] and -r [ɻ], are both represented as [l], the only approximant in Korean. Table 2.30  Korean transcription of Mandarin approximants Pinyin

IPA

Character

Hangul

IPA

gélì tóngréntáng

kɤ.li tʰuŋ. ɻən.tʰɑŋ

格力 同仁堂

거리 퉁런탕

kʌ.li tʰuŋ.lʌn.tʰaŋ

2.3.1.1.5 NASALS

The following chart lists the nasals in Mandarin and Korean. Table 2.31  Nasals in Mandarin and Korean

Mandarin Korean

Bilabial

Alveolar

Velar

m m

n n

ŋ ŋ

Mandarin 41 The Mandarin [m] and [n] correspond to [m] and [n] in Korean, respectively. The [ŋ] cannot appear in the initial position in both Mandarin and Korean. The Mandarin [ŋ] in the coda position is represented as [ŋ] in Korean. Table 2.32  Korean transcription of Mandarin nasals Pinyin

IPA

Character

Hangul

IPA

mǔdānjiāng nánjīng

mu.tan.tɕjaŋ nan. tɕiŋ

牡丹江 南京

무단장 난징

mu.tan.tɕaŋ nan.tɕiŋ

2.3.1.1.6 SUMMARY

Based on our above discussion, the following table summarizes the Korean government transcription of Mandarin consonants in Korean. Table 2.33  Korean government transcription of Mandarin consonants

Stop

voiceless unaspirated voiceless aspirated

Fricative

Affricative

dental-alveolar retroflex palatal

Nasal Approximant Lateral

bilabial dental-alveolar velar

Pinyin

Mandarin consonants (IPA)

Korean consonants (IPA&Hangul)

b d g p t k f s sh x h z c zh ch j q m n ng r l

p t k pʰ tʰ kʰ f s ʂ ɕ x ts tsʰ tʂ tʂʰ tɕ tɕʰ m n ŋ ɻ l

pㅂ tㄷ kㄱ pʰ ㅍ tʰ ㅌ kʰ ㅋ pʰ ㅍ s*ㅆ sㅅ sㅅ hㅎ tɕ* ㅉ tɕʰ ㅊ tɕ ㅈ tɕʰ ㅊ tɕ ㅈ tɕʰ ㅊ mㅁ nㄴ ŋㅇ lㄹ lㄹ

2.3.1.2 Vowels 2.3.1.2.1  HIGH VOWEL

The following chart lists the high vowels in Mandarin and Korean.

42  Mandarin Table 2.34  High vowels in Mandarin and Korean Front High

unrounded i, ɿ iㅣ

Mandarin Korean

Central rounded y yㅟ

Back unrounded ʅ ɯㅡ

rounded u uㅜ

The high vowel [i] and [u] exist in both Mandarin and Korean, so a perfect mapping is seen in these two vowels. Following the prescription of SKP, the Mandarin high vowel [y] has a corresponding phoneme [y] 위 in Korean. However, native speakers of Seoul Korean usually pronounce the vowel [y] ㅟ as a diphthong instead of a monophthong (IPA handbook 1999; Lee and Ramsey 2000; Shin et al. 2012). Nevertheless, Mandarin high vowel [y] is transcribed as Korean [y] 위 in the Korean government transcription. Table 2.35  Korean transcription of Mandarin high vowels [i] and [y] Pinyin

IPA

Character

Hangul

IPA

yìwū yùshù

i.u y.ʂu

義烏 玉樹

이우 위수

i.u y.su

The Mandarin [ɿ] and [ʅ] are not seen in the vowel inventory of Korean. The Korean [ɯ], an unrounded back vowel, is used to represent both [ɿ] and [ʅ]. Table 2.36  Transcription of Mandarin high vowels [ɿ] and [ʅ] in Korean Pinyin

IPA

Character

Hangul

IPA

zībó chángzhì

tsɿ.po tʂʰɑŋ.tʂʅ

淄博 長治

쯔보 창즈

tɕ*ɯ.po tɕʰaŋ.tɕɯ

2.3.1.2.2  MID VOWEL

The following chart lists the mid vowels in Mandarin and Korean. Table 2.37  Mid vowels in Mandarin and Korean Front Mid

Mandarin Korean

unrounded ɛ    e eㅔæㅐ

rounded øㅚ

Central

Back

ə

unrounded ɤ ʌㅓ

rounded o oㅗ

The mid vowel [e] and the vowel [o] exist in both Mandarin and Korean, so there is a perfect match in these vowels.

Mandarin 43 Table 2.38  Korean transcription of Mandarin mid vowels [e] and [o] Pinyin

IPA

Character

Hangul

IPA

hēilóngjiāng zībó

xei. luŋ. tɕjaŋ tsɿ.po

黑龍江 淄博

헤이룽장 쯔보

he.i.luŋ.tɕaŋ tɕ*ɯ.po

The Mandarin vowel [ɛ] is represented as [e] in Korean, as shown below. Table 2.39  Korean transcription of Mandarin mid vowel [ɛ] Pinyin

IPA

Character

Hangul

IPA

tiānjīn

tʰjɛn. tɕin

天津

톈진

tʰjen.tɕin

The Mandarin vowel [ə] and [ɤ] are both represented as [ʌ], a back vowel in Korean. Table 2.40  Korean transcription of Mandarin mid vowels [ə] and [ɤ] Pinyin

IPA

Character

Hangul

IPA

xiàmén gélì

ɕja. mən kɤ.li

廈門 格力

샤먼 거리

ɕja.mʌn kʌ.li

2.3.1.2.3  LOW VOWEL

The following table lists the low vowels in Mandarin and Korean. Table 2.41  Low vowels in Mandarin and Korean Front Low

Mandarin Korean

Central

unrounded rounded

a aㅏ

Back unrounded rounded

As shown in Table 2.41, the [a] vowel exists in both Mandarin and Korean. The Mandarin vowel [a] is faithfully adapted into Korean [a]. Table 2.42  Korean transcription of Mandarin low vowel Pinyin

IPA

Character

Hangul

IPA

zhāngjiājiè

tʂaŋ. tɕja. tɕje

張家界

장자제

tɕaŋ.tɕa.tɕe

2.3.1.2.4 SUMMARY

Based on our above discussion, the following table summarizes the adaptation of the Mandarin vowels in Korean.

44  Mandarin Table 2.43  Korean transcription of Mandarin vowels Pinyin

Mandarin vowel (IPA)

Korean vowel (IPA)

i ü i i u e e e o a

i y ɿ ʅ u ɤ e ə o a

i y ɯ ɯ u ʌ e ʌ o a

2.3.1.3  Syllable structure The structure of a Mandarin syllable is (C)(G)V(C/V), while the structure of a Korean syllable is (C)(G)V(C). The following syllable types in Mandarin are mapped into one syllable in Korean in the government transcription. Table 2.44  Syllable correspondence between Mandarin and Korean (1)

V CV GV CGV VC CVC GVC CGVC

Pinyin

IPA

Character

Meaning

Korean Transcription

IPA

è mǎ yā duō ān lán yán huàn

[ɤ]51 [ma]214 [ja]55 [two]55 [an]55 [lan]35 [jɛn]35 [xwan]51

餓 馬 鴨 多 安 藍 鹽 換

hungry horse duck many peace, safe blue salt to exchange

어 마 야 둬 안 란 옌 환

ʌ ma ja twʌ an lan jen hwan

However, the Mandarin (V) at the end of a diphthong or a triphthong, which is not allowed in a Korean syllable, is represented as an independent syllable. Table 2.45  Syllable correspondence between Mandarin and Korean (2)

VV CVV GVV CGVV

Pinyin

IPA

Character

Meaning

Korean Transcription

IPA

ài lái yào huài

[ai]51 [lai]35 [jɑu]51 [xwai]51

愛 來 藥 壞

love come medicine bad

아이 라이 야오 화이

a.i la.i ja.o hwa.i

Following the Korean government transcription methods, the [j] glide in Mandarin is dropped in the Korean representation when the preceding Mandarin initials are adapted in Korean as ㅈ[tɕ], ㅉ[tɕ*], or ㅊ[tɕʰ].

Mandarin 45 Table 2.46  Korean transcription of Mandarin [j] glide Pinyin

IPA

Character

Hangul

IPA

zhāngjiājiè qiónglái

tʂaŋ. tɕja. tɕje tɕʰjuŋ. lai

張家界 邛崍

장자제 충라이

tɕaŋ.tɕa.tɕe tɕʰuŋ.la.i

In addition, the Mandarin -iu, -ui and -un are phonetically [jou], [wei] and [wən], although the nucleus is not spelled out in Hanyu pinyin and Wade-Giles. Influenced by the orthography, the nucleus of Mandarin -ui [wei] and -un [wən] is dropped in the Korean representation. Table 2.47  Korean transcription of Mandarin [wei] and [wən] Pinyin

IPA

Character

Hangul

IPA

ānhuī chángchūn

an.xwei tʂʰaŋ.tʂʰwən

安徽 長春

안후이 창춘

an.hu.i tɕʰaŋ.tɕʰun

As for the Mandarin final -iu [jou] there are two types of representations in Korean. First, when the initial is not -j [tɕ] or -q [tɕʰ], the nucleus in the -iu [jou] is dropped in the Korean representation. Table 2.48  Korean transcription of Mandarin [jou] (1) Pinyin

IPA

Character

Hangul

IPA

liúyáng

ljou. jaŋ

瀏陽

류양

lju.jaŋ

Second, when the initial is -j [tɕ] or -q [tɕʰ], both the glide and the nucleus in the -iu [jou] are dropped. Table 2.49   Korean transcription of Mandarin [jou] (2) Pinyin

IPA

Character

Hangul

IPA

jiǔjiāng shāngqiū

tɕjou.tɕjaŋ ʂɑŋ.tɕʰjou

九江 商丘

주장 상추

tɕu.tɕaŋ saŋ.tɕʰu

2.3.1.4 Tones According to the Korean government transcription, Mandarin tones are not transcribed into Korean, as shown below. Table 2.50  Korean transcription of Mandarin tones Pinyin

IPA

Character

Hangul

IPA

wúxī wūxī

u.ɕi u.ɕi

无锡 巫溪

우시 우시

u.si u.si

46  Mandarin Tone does not exist in the phonological system of Standard Korean, and no super-segmental feature in Standard Korean is available for the adaptation of Mandarin tones. Therefore, Mandarin tones are not transcribed into Korean. It is noteworthy that Ito and Kenstowicz (2009) found that in Yanbian Korean, which has a pitch-accent system, faithfulness to the first mora of the final syllable plays a crucial role in the adaptation of Mandarin tones into the pitch pattern in Yanbian Korean. 2.3.2  Choe-Kim (1985) One year before the announcement of the South Korean government transcription system (KGT), Choe and Kim put forward their proposal for the transcription of Mandarin loanwords in Kim (1985). Table 2.51 summarizes the differences between the government’s and Choe and Kim’s (1985) transcription systems. In this table, a dot (.) indicates a syllable boundary, and parentheses indicate the alternative representation when finals are preceded by consonant initials. Table 2.51  Differences between Choe-Kim’s (1985) transcription and KGT Mandarin Initial

Final

Pinyin b d g ch f r o ong ia ie iu iong ian iang ua uo ui uan uang üe üan

IPA p t k tʂʰ f ɻ o uŋ ja jɛ jou juŋ jɛn jɑŋ wa wo wei wan waŋ ɥɛ ɥɛn

Government

Choe-Kim

IPA p t k tɕʰ ph l o uŋ jɑ je ju juŋ jen jɑŋ wɑ wʌ we.i (u.i) wɑn wɑŋ we y.ɑn

IPA p* t* k* tɕʰɯ hw lɯ wʌ oŋ jɑ (i.ɑ) je (i.e) ju (i.u) joŋ (i.oŋ) jen (i.en) jɑŋ (i.ɑŋ) wɑ (u.ɑ) wʌ (u.ʌ) we.i wɑn (u.ɑn) wɑŋ (u.ɑŋ) y.e y.æn

The primary intention of Choe-Kim’s (1985) proposal is to present a system of transcription that maintains the original distinctions of Mandarin phonemes. For the Mandarin phonemes that Korean has no corresponding sounds, ChoeKim (1985) proposed to use special digraphic spellings to distinguish them. For instance, Mandarin [x] is transcribed as [h], and Mandarin [f], which has no

Mandarin 47 corresponding sound in Korean, is transcribed as [hw]. Mandarin [l] is transcribed as [l], and Mandarin [ɻ], a sound not present in Korean, is transcribed as [lɯ]. In order to keep the distinctiveness in the source form, Choe-Kim’s (1985) proposal sacrificed the faithfulness to the original sound. This is different from the normal loanword borrowing process, the goal of which is pursing closeness to the original form within the phonemic inventory of the borrowing language. Another feature of Choe-Kim’s (1985) transcription is to maximize the perceptual similarity. For example, the Mandarin -b [p], -d [t] and -g [k], which are transcribed as Korean lax stops ㅂ[p], ㄷ[t] and ㄱ[k] in the government transliteration system, are transcribed as Korean tense stops [p*], [t*] and [k*] in Choe-Kim’s (1985) system owning to the situation that Mandarin -b [p], -d [t] and -g [k] are perceptually more similar to Korean tense stops. The government transcription of Mandarin [ɥɛn] is [y.an], which is probably influenced by the pinyin spelling -üan (uan). To reflect the pronunciation in actual speech, Mandarin [ɥɛn] was transcribed as [y.æn] in Choe-Kim (1985). 2.3.3  Eom (1996) Eom (1996) proposes a series of revisions to the government transcription. Below is a summary of the difference between the Korean government transcription and Eom’s (1996) transcription. 2.3.3.1 Consonant Table 2.52  Eom’s (1996) transcription of Mandarin consonants [p], [t] and [k] Mandarin voiceless unaspirated stop

Pinyin b d g

Government IPA p t k

Hangul ㅂ ㄷ ㄱ

Eom IPA p t k

Hangul ㅃ ㄸ ㄲ

IPA p* t* k*

As mentioned above, although the Mandarin -b [p], -d [t], -g [k] and the Korean ㅂ[p], ㄷ[t], ㄱ[k] are both transcribed as [p, t, k] in IPA, many studies show that the Mandarin -b [p], -d [t] and -g [k] is perceived to be more similar to the Korean tense stops [p*], [t*] and [k*] than the lax consonants ㅂ[p], ㄷ[t], and ㄱ[k] (Eom 1996; Jeon 1999). Therefore, Mandarin -b [p], -d [t] and -g [k], which are transcribed as Korean lax stops ㅂ[p], ㄷ[t] and ㄱ[k] in the government transliteration rule, are transcribed as Korean tense stops [p*], [t*] and [k*] in Eom (1996) based on perceptional similarity. 2.3.3.2 Glide The creators of the government transcription system believe that the [j] glide after the [tɕ] and [tɕh] is not necessary, since [tɕ] and [tɕh] are palatalized initials. On

48  Mandarin the contrary, [j] glide is still written out in Eom’s (1996) transcription system. Eom (1996: 69) argues that the regulation to drop the [j] glide after Mandarin [tɕ] or [tɕh] in the government transcription system causes two different phonemes in the source language to be represented in an identical way, and therefore should be repealed. For example, both Mandarin [tɕja], [tɕhja], [tɕjɑu], [tɕhjɑu], [tɕjɑŋ], [tɕhjɑŋ] and Mandarin [tʂɑ], [tʂʰɑ], [tʂɑu], [tʂʰɑu], [tʂɑŋ], [tʂʰɑŋ] are adapted as [tɕɑ], [tɕhɑ], [tɕɑ.o], [tɕhɑ.o], [tɕɑŋ] and [tɕhɑŋ] in the government transcription. 2.3.3.3 Vowel Table 2.53  Eom’s (1996) transcription of Mandarin vowels Mandarin

Government

Eom (1996)

Mandarin

Government

Eom (1996)

ɑu ou ja jɛ jɑu jou jɛn jɑŋ

ɑ.o ʌ.u jɑ je jɑ.o ju jen jɑŋ

ɑ.u o.u jɑ (i.ɑ) jæ (i.æ) jɑ.u jo.u (i.u) jæn (i.æn) jɑŋ (i.ɑŋ)

juŋ wa wo wei wan waŋ ɥɛ ɥɛn

juŋ wɑ wʌ we.i (u.i) wɑn wɑŋ we y.ɑn

juŋ (i.uŋ) wɑ (u.ɑ) wʌ (u.o) we.i wɑn (u.ɑn) wɑŋ (u.ɑŋ) y.æ y.æn

This revision in Table 2.53, according to Eom (1996: 73), is made to maximize the perceptual similarity to the source input. The government transcription of Mandarin -uan [ɥɛn] is [y.an]. In fact, the nuclear vowel in Mandarin -uan is [ɛ] in actual speech, although it is written as -a in Hanyu pinyin. In order to be faithful to the source input form, Eom (1996) transcribes Mandarin -uan [ɥɛn] as [y.æn]. We can notice from Table 2.53 that some monosyllabic representations in the government transcription are revised into disyllabic forms when the final is preceded by a consonant initial. Eom (1996: 77) claims that the government transcription includes many monosyllabic representations which are rarely used in native Korean phonology, such as [pje], [mjen], [thwɑn] etc. In Eom (1996) these syllable types are avoided by appealing to the resyllabification strategy. 2.3.4  Eom (2002) Eom (2002) revised Eom’s (1996) transcription of those finals that are listed in Table 2.54. The principle of the revision is to rank the faithfulness to Hanyu pinyin spelling above the maximal preservation of the perceptual similarity to the source input (Eom 2002: 87). Eom (2002) revises the transcription of Mandarin [ɥɛ] from [y.æ] to [y.e] when preceded by a consonant initial and [we] elsewhere. The monosyllabic form [we]

Mandarin 49 Table 2.54  Eom’s (2002) transcription of Mandarin vowels Mandarin Pinyin ao ong iao iong ie üe wei (ui)

IPA ɑu uŋ jɑu juŋ jɛ ɥɛ wei

Eom (1996)

Eom (2002)

IPA ɑ.u uŋ iɑ.u juŋ (i.uŋ) jæ (i.æ) y.æ we.i

IPA ɑ.o oŋ iɑ.o joŋ (i.oŋ) je (i.e) we (y.e) we.i (u.i)

is proposed to pursue the simplicity of representation (Eom 2002: 91). Given phonotactic constraints in native Korean phonology, the combination of [y] and [e] within one syllable, namely [ye], is not allowed; thus, [we] becomes the output form.

2.4  Loanword phonology Loanword phonology is a subfield of phonology which investigates the mapping between two phonological systems in the process of borrowing. Loanwords show the conflicts between the sound systems of two languages. On the one hand, loanwords preserve a source language as much as possible; on the other hand, they have to conform to the target language’s phonological system. The mechanisms and processes by which loanwords are adapted and transmitted from a language to another is the main interest of loanword phonology studies. The main questions that loanword phonology addresses include the following: One major question is what is the original input form that entered the process of adaptation. Is it the acoustic form that is perceived by the borrower, or the phonological form that is filtered, according to phonological constraints in the target language, or the form reflected in the orthography of the source language? Another major question is how the input is adapted to fit the phonological system of the target language. The sounds or phonological structures that are not available in the target language will be borrowed in the most faithful form possible. This will involve certain constraints and objectives, such as the wellformedness restriction and the goal of maximal preservation. In consideration of these processes, one major question is what is the interaction between the well-formedness restriction of the target language and the maximal preservation of the input form. Scholars have proposed many different views on what are the major patterns of adaptation, and how they should be interpreted. The majority of the theory of loanword phonology can be condensed into three main approaches, that is, the perception approach, phonology approach, and the phonetics-phonology approach (Lin

50  Mandarin 2009; Kim 2012; Yu 2014). Apart from these three major approaches, many scholars (e.g., Miao 2005; Smith 2006; Vendelin and Peperkamp 2006) have found that a variety of other factors, including orthography, morphology and semantics, have also been reported in a number of studies as having effects on loanword phonology. 2.4.1  Perception approach The perception approach claims that the perception of the borrower determines the result of borrowing in the loanword adaptation (Dupoux et al. 1999; Peperkamp 2005; Peperkamp and Dupoux 2003; Vendelin and Peperkamp 2004; Peperkamp, Vendelin and Nakamura 2008). Proponents of this approach argue that the borrower does not access the phonology of the source language, but the inputs of loanwords are on the level of surface forms. Dupoux et al. (1999) reported that when native speakers of Japanese are presented with the form [ebzo], they report that what they hear is [ebɯzo] in most cases. Moreover, Japanese participants have severe difficulties to distinguish between [ebzo] and [ebɯzo] in discrimination tasks. Peperkamp and Dupoux (2003) compared loanword adaptations to experimental speech perception data and found many correspondences. For instance, Korean listeners find it difficult to distinguish [ɹ] and [l] and in loanwords originating from English, with wordinitial -l often being adapted as [r] in Korean; likewise, French listeners have difficulties in perceiving stress contrasts, and foreign loanwords in French, therefore, invariably have word final stress, regardless of the stress in the original source form. Accounting for these linguistic facts, Peperkamp and Dupoux (2003: 367) proposed that “loanword adaptation take place in perception and are defined as phonetically minimal transformations.” Peperkamp (2005) pointed out that conflicts between native phonological alternation and loanword adaptations as well as “unnecessary” adaptation are highly problematic for analyses that derive adaptations within the phonological grammar of the borrowing language. For instance, [s] is not allowed in a syllable coda in the native Korean phonology. An underlying /s/ is realized as [t] when it occurs in coda position (e.g. 맛 “taste” /mas/[mat]), but in loanwords, words with [s] in a coda position will undergo a process of epenthesis. Also, in Korean, loanwords from English with a voiceless stop coda are often adapted with an aspirated stop followed by an epenthetic vowel, especially when the preceding vowel is tense, such as bat > [pæ.thɯ], deck > [tɛ.khɯ], hip > [hi.ph]. These adapted forms are unexpected, since the voiceless stop coda can appear in the coda position in native Korean phonology, such as [pat] “field,” [kæk] “guest” [ʨip] “house.” Peperkamp (2005: 10) concludes that these two phenomena can be explained “under the hypothesis that loanword adaptations are basically phonetic rather than phonological in nature.” Vendelin and Peperkamp (2004) also pointed out that in Japanese loanwords, word-final [n] is adapted as a moraic nasal consonant in loanwords from English, whereas a geminate nasal followed by an epenthetic vowel in loanwords from French. In order to explain this phenomenon, Vendelin and Peperkamp (2004)

Mandarin 51 conducted a forced-choice task experiment and found that Japanese listeners perceived epenthetic vowels significantly more often in stimuli produced by French speakers than that produced by American English speakers. Peperkamp, Vendelin and Nakamura (2008) also examined the different adaptation of word-final [n] in Japanese loanwords from English and French through a related experiment and found that French word-final [n] has a strong vocalic release, which is perceived by Japanese listeners as Japanese native vowel [ɯ]. The perception approach is criticized in the literature for the following reasons. First, since language borrowing is usually first exhibited by competent bilinguals, it is doubtful that the borrower truly has no knowledge of the source language phonology (Paradis 1996; LaCharité and Paradis 2005). Second, many studies show that perception is not the only factor which is productive in loan adaptation, even though it has been clearly attested to be an important factor (Takagi and Mann 1994; Smith 2006). Third, many counter examples can be put forward in direct opposition to the perception approach. For instance, the English word “tie” and “stick” are rendered as [tʰai] and [sɯ.tʰik] in Korean. However, the /t/ in the English word “stick” is perceptually more similar to the tense plosive [t*] in Korean. In other words, perception was clearly not the only fact in the borrowing of these words. 2.4.2  Phonology approach The phonology approach claims that loanword adaptation is a process primarily involving phonological processes (Itô and Mester 1995; Paradis 1996; Paradis and LaCharité 1997; Jacobs and Gussenhoven 2000; LaCharité and Paradis 2005, etc.). In this approach, it was argued that linguistic inputs which entered a process of loanword adaptation reflect the phonological categories of the source language. Smith (2006) proposed that the phonological system of the target language also plays an important role in loanword adaptation. The loanword from the source language is adapted into the form that fit the phonological requirements of the target language. According to Smith (2006), the English word jitterbug is adapted as jittaabaggu in the orthographic borrowing in Japanese. Smith claims that the [ɯ] in the jittaabaggu [ʤit.ta:.bag.gɯ] is not provided in the English orthography, so it is the Japanese phonotactic constraint that derives the Japanese SR [ʤit.ta:. bag.gɯ] by the insertion of an epenthetic vowel. LaCharité and Paradis (2005) provided several examples to support the phonology approach. The first example is from the borrowing of English obstruents in Mexican Spanish. In terms of VOT, the English voiced stops should be matched to Spanish voiceless stops given that both sounds have a VOT of 0–30 milliseconds. However, English loanwords with voiced stops are adapted into [voiced] types in Spanish, which is phonologically more similar. In addition, the adaptation of the English high lax vowels in French also follows a rule of category proximity, rather than phonetic proximity. Paradis (1996) and Paradis and LaCharité (1997) proposed the theory of constraints and repair strategies (abbreviated TCRS) to account for adjustment during

52  Mandarin adaptation. According to this theory, the variation between segmental preservation and deletion is restricted depending on how many repair strategies are used. This theory argues that once preservation of a segment exceeds an acceptable threshold, deletion will occur. For example, in “avocat” [avɔka] (French for lawyer)>[awɔka] (Fula), the preservation of /v/ by /v/ > /w/ mapping requires a one-step repair. That is the insertion of a [+son] feature. On the contrary, deletion is adopted in “voyage” [vwajaʒ] (French) > [waja:s] (*[wuwaja:s]) (Fula) because preservation would involve a three-step rescue. They are the insertion of a nucleus node, spreading of the glide [w] to the nucleus position, and adaptation of nonnative [v] into [w] in native language. Other studies propose to explain the adaptation of loanwords from a constraintranking optimality theoretic (abbreviated OT) perspective. Itô and Mester (1995) argued that the re-ranking of the same set of faithfulness constraints can explain differences in the strata of the Japanese lexicon, explaining for a gradual degradation, that is, a “weakening” of the “force” of various well-formedness constraints. Davidson and Noyer (1997) proposed an analysis of the stress assignment in Hauve adaptation of Spanish loanwords by re-ranking faithfulness constraints. Jacobs and Gussenhoven (2000) also argued that loan phonology is not a separate component of the grammar and that phonological adjustments in the process of borrowing must be accounted for by the same constraint hierarchy that characterizes native phonology. However, one major problem with the phonology approach is its failure to explain all the patterns in the loanword adaptation, which was the cause of much criticism. For instance, the English words “tie” and “stick” are borrowed as [tʰai] and [si.tik] in Cantonese (Silverman 1992). The aspiration feature of these two words is faithfully reflected in Cantonese. This argues against the phonology approach which claims that the non-phonemic features in the source language will not be adapted in the target language. 2.4.3  Phonetics-phonology approach The phonology approach can explain loanword adaptation via phonological process, but it overlooks the role of perception. On the other hand, the perception approach ignores the fact that other factors, such as lexical, phonological, and orthographic knowledge of the source language could affect the loanword adaptation. The phonetics-phonology approach argues that inputs in the process of adaptation are based on how the borrowing language perceives acoustic information in the source language, and the inputs next are modified for adaptation based on perception by the phonological restrictions of the target language (Silverman 1992; Yip 1993, 2002, 2006; Kang 2003; Kenstowicz 2003; Iverson and Lee 2006; Lin 2009). This approach involves both perception and phonology and treats them as separate levels that are governed by different grammatical systems (Silverman 1992; Yip 1993; Kenstowicz 2003). According to this theory, the recipient language speakers have no access to the original input: the input which speakers will access and process in borrowing and adaptation is the output

Mandarin 53 of the perception level. In other words, it is a form modified from the source pronunciation. Silverman (1992) claims there are two steps in the process of loanword adaptation, namely the “perceptual level” and “operative level.” On the perceptual level, inputs are acoustic signals. Therefore, segments which are not salient in acoustic signals will be ignored. Whether or not segments are filtered in the “perceptual level” is determined by a degree of saliency and if the segments exist in the phonological inventory of the borrowing language. Furthermore, segments which already passed “perceptual level” are adjusted based on the phonotactic constraints of the borrowing language. For instance, the unaspirated allophone of an English voiceless stop — such as [t] in “stick” — is perceived as an unaspirated stop in Cantonese in the “perception level”; next at the “operative level” native phonotactic constraints adjust the perceived sound to fit native syllable structures. In other words, English “stick” is perceived as /stik[H]/ at the perceptual level, and adapted as /si[L] tik[H]/ at the operative level to represent the consonant cluster. Yip (1993) adopts Silverman’s (1992) concept of the perception and operation levels. However, Yip’s (1993) analysis on the operation level is within the framework of OT, and she claims that phonological constraints on loan phonology and native phonology are actually identical. In another study by Kenstowisz (2003), it is argued that in the Fon adaptation of French loanwords, the perception level and the operation level have different constraint hierarchies. In the perception level, the constraint against vowel insertion (i.e. Dep-V) has higher priority than the constraint against consonant deletion (i.e. Max-C). For example, “poste [post]” is adjusted as [pos] in the perception level. However, in the operation level, the Max-C constraint has the priority over Dep-V. Consequently, [pos] in the perception is readjusted as [posu].

2.5 Discussion 2.5.1  Reflection on perception and phonology approaches The perception approach and the phonology approach alone cannot explain all Mandarin loanwords in Korean. First, the perception approach claims that the mismatch between the source language and the adapted form in the target language is caused by the misperception of borrowing forms. In other words, the borrower cannot perceive the source form accurately. However, many counter examples to this argument can be found in various examples of Mandarin loanwords and hangul transcription systems in Korean. First, one example is that Mandarin -ui [uei] is represented as 우이 [u.i] in Korean, with the nucleus [e] in Mandarin being deleted in its Korean adaptation. At first glance, this appears to be a deletion of the nucleus [e] caused by the borrower failing to perceive this segment due to its phonetic weakness. However, although the official document Oeraeeo pyogibeop 외래어 표기법 (Rules for writing foreign language words in Hangul) states that the Mandarin final -ui [uei] is represented as the form 우이

54  Mandarin [u.i], next to the word 우이 [u.i], the form 웨이 [we.i], which retains the nucleus [e], is also given in parentheses. Therefore, we can conclude that the nucleus [e] can be perceived by the borrower, yet it is still deleted in the representation. In other words, it is clear that other factors — specifically in this particular example, Mandarin’s orthographic representation — are productive factors in adaptation. Second, another example is the following: many studies show that the Mandarin -b [p], -d [t] and -g [k] are perceived to be much closer to the Korean tense stops [p*], [t*] and [k*] than the lax consonants ㅂ[p], ㄷ[t] and ㄱ[k] (Eom 1996; Jeon 1999); however, despite their perceptional similarity with Korean tense stops, the Mandarin -b [p], -d [t] and -g [k] are nonetheless transcribed as lax consonants ㅂ[p], ㄷ[t] and ㄱ[k] in the Korean government transcription. A counter example to the phonology approach can be found in the adaptation of Mandarin fricative -s [s] in Korean. The Mandarin fricative -s [s] is adapted as the tense consonant [s*] in Korean. Although Mandarin fricative [s] and Korean lax consonant [s] belong to the same phonological category, the Mandarin fricative -s [s] is not represented as the lax [s] in Korean transcription systems, but rather the tense fricative [s*], which is a perceptually similar phoneme. The above discussion shows that in order to account for the Mandarin loanwords in Korean, both the perception and phonology should be considered. 2.5.2  Influence of orthography We can find the influence of orthography in all the transcription systems introduced above. The adaptation of Mandarin -iu [jou] and -un [wən] in ChoeKim (1985), the government transcription and Eom (1996, 2002) can be easily explained through the influence of orthography, as shown in Table 2.55. Table 2.55  Influence of orthography on the transcription Mandarin Pinyin diu jiu liu dun gun

Choe-Kim IPA tjou tɕjou ljou twən kwən

Hangul 띠우 지우 리우 뚠 꾼

Government IPA t*i.u tɕi.u li.u t*un k*un

Hangul 듀 주 류 둔 군

Eom IPA tju tɕu lju tun kun

Hangul 띠우 지우 리우 뚠 꾼

IPA t*i.u tɕi.u li.u t*un k*un

The influence of orthography is also clearly discernible in the Korean government transcription of Mandarin [ɥɛn] as [y.ɑn]. The nuclear vowel of [ɥɛn] is written as -a in both Hanyu pinyin and Wade-Giles, even though the phonetic value of this nuclear vowel is regularly [ɛ] in actual speech in Mandarin. 2.5.3  Role of the borrower The situation where borrowers with the same native language may propose different adapted forms for the same source input can be attributed to the following two factors: the inventory of constraints and the ranking of constraints.

Mandarin 55 2.5.3.1  Inventory of constraints The proposed constraints in the government transcription and Eom (2002) differ in the following aspects. First, despite perceptional similarity with Korean tense stops, the Mandarin -b [p], -d [t] and -g [k] are transcribed as Korean lax stops ㅂ [p], ㄷ[t] and ㄱ[k] in the Korean government transcription. This can be explained by a restriction in the government transcription rule, which states that “a plosive should not be represented by a tense consonant.” This constraint, however, is not seen in Choe-Kim (1985) and Eom (2002). Based on perceptional similarity, Choe-Kim (1985) and Eom (2002) transcribe Mandarin -b [p], -d [t] and -g [k] as Korean tense stops [p*], [t*] and [k*]. Second, the [j] glide after [tɕ] or [tɕh], which is dropped in the government transcription, is preserved in Choe-Kim (1985) and Eom (2002), such as [tɕjau] > [tɕa.o] (government)/[tɕja.o] (Choe-Kim 1985, Eom 2002). The creators of the government transcription rules propose a constraint to prevent the occurrence of -j glide after -tɕ or -tɕh. This constraint, however, is not present in Choe-Kim (1985) and Eom (2002). Third, diphthongs and triphthongs in Mandarin, which are represented as one syllable in Korean transcription systems, except for -ai, -ei, -ao and -ou in the government transcription, are uniformly represented as two syllables in Eom (2002). In order to maintain the single-syllable representation, the government transcription includes many syllable types which are observed in onomatopes, such as [twan] 돤, [tɕhwaŋ] 촹 etc. Eom’s transcription, however, avoids the use of these syllable types by adopting the resyllabification strategy. The constraint on those low-frequency syllable types, which is seen in Eom (2002), does not exist in the government transcription. 2.5.3.2  Ranking of constraints With the same number of constraints, the ranking of these constraints could vary for the following three reasons. First, different borrowers may propose different rankings. For instance, the -o in -ong is represented as [uŋ] in the government transcription, while it is represented as [oŋ] in Choe-Kim (1985) and Eom (2002). In this case, perceptional similarity is prioritized over source form spelling in the government transcription, while source form spelling is prioritized over perceptional similarity in Choe-Kim (1985) and Eom (2002). Second, the ranking of constraints proposed by the same borrower may also be subject to change. For example, Eom (1996) proposes to transcribe Mandarin -ong as [uŋ] out of the consideration for its perceptional similarity to the source form. However, the -ong is transcribed as [oŋ] in Eom (2002). Eom (2002: 87) writes that “the principle of convenience (having transcription directly based on Hanyu pinyin) is given superiority to the principle of faithfulness to the source form.” Third, the same translator may propose different ranking in different words. Loanword borrowing even when observed by the same borrower could present different rankings of constraints and consequently obtain different results. For example, the -o in -ao and -o in -ong in Mandarin are both [u]. However, in the government transcription,

56  Mandarin the former is represented as [o] in [a.o], whereas the latter is represented as [u] in [uŋ]. Therefore, in such a case, consistency to pinyin spelling is given priority to the perceptional similarity in the case of -ao, but the ranking of constraints is reversed in the case of -ong. 2.5.4 Summary The result of this case study on the adaptation of Mandarin loanwords in Korean supports the phonetics-phonology approach in loanword phonology. In other words, both the perception and phonology play an important role in the process of adaptation. Previous studies in loanword phonology aim to find one single theory to explain all adapted forms in the same language. However, it is hard to find a clear-cut pattern with a single adaptation approach. This study demonstrates that it is necessary for researchers to pay attention to the role of the borrowers because they may adopt different strategies, involving different constraints as well as different rankings or priority of constraints, in the process of borrowing. Different borrowers may also have different attitudes toward the marked form and have different understandings of the native phonology. What is the best or the most faithful way to transcribe loanwords could be different from one speaker’s mind to another’s. And even for the same form in the source language, the same borrower may propose different adapted forms for different words.

Notes 1 There are also syllabic consonants in Mandarin, such as answering words [m̩ ] 呣 and [n̩ ] 嗯. 2 The symbol before the Hangul alphabet is the Yale Romanization of Korean. The complete list of the Yale Romanization of Korean alphabet is shown as follows: ㅂ -p, ㅃ-pp, ㅍ-ph, ㅁ -m, ㄷ-t, ㄸ-tt, ㅌ-th, ㄴ-n, ㄹ -l, ㅈ -c, ㅉ-cc, ㅊ-ch, ㄱ-k, ㄲ-kk, ㅋ-kh, ㅇ-ng, ㅅ -s, ㅆ -ss, ㅎ-h. 3 The Yale Romanization of Korean vowels are listed as follows: ㅣ-i, ㅟ -wi, ㅡ-u, ㅜ-wu, ㅔ-ey, ㅚ -oy, ㅓ-e, ㅗ-o, ㅐ-ay, ㅏ -a, ㅑ-ya, ㅛ-ye, ㅛ-yo, ㅠ-yu, ㅒ-yay, ㅖ-yey, ㅘ-wa, 궈 -we, ㅙ -way, ㅞ -wey, ㅢ -uy. 4 [ŋ] cannot appear at the beginning of a word. It can appear phonetically at the beginning of a syllable only if the syllable is non-initial in a word. 5 This document is available on the webpage of National Institute of the Korean Language: www.korean.go.kr/09_new/dic/rule/rule_foreign_index.jsp (accessed 10 January 2015) 6 The pronunciation of ㅅ is [s] in the initial position and [t] in the coda position.

3 Initial system of Middle Sino-Korean

This chapter will compare the initial system of Middle Chinese with that of EMSK. The initial system of MC underwent a change between the time of Qieyun and recording of the rhyme tables produced in the Song dynasty. The initial system of LMC differs from Qieyun mainly in the following two aspects: First, LMC makes distinction of bilabial bang 幫 series and labial-dental fei 非 series; second, the xia 匣 initial before the third division becomes the yun 云 initial in LMC. The initial system of EMSK introduced in this study is based on source data found in HMJH, CJM and YH. Phonetic notation for 4,803 characters, in total, can be found in these three manuscripts. It is important to note that those characters that have more than one phonetic notation are counted as individual characters. Where the notation for the same character is not consistent in these three books, only the notation provided from HMJH will be provided in the comparison table, and the notations in the other two sources will be given in the footnote. When the notation for the same character differs between CJM and YH, the notation given in YH will be provided in the comparison table, and the notation in CJM will be given in the footnote. It should be noted that the transcription conventions in these three books differs with respect to the transcription of the ri 日 initial. In order to avoid confusion, the reading that found from CJM is underlined with a wave line and that from YH with a straight line where necessary. In order to provide a more comprehensive understanding of the initial system in MSK, for those non-mainstream representations, the phonetic notations in all the sources consulted will be provided in the footnote. The reading of Chinese characters will be arranged into tables according to their initials in MC. In each chart, rhyme information, kaihe “labial roundness” and deng “division” as well as the phonetic reconstruction in Pan and Zhang (2013) will be provided for reference. At the end of each chart, one can find the number of occurrences for each type of representation in the observed data.

3.1 Group bang 幫 (bilabials) The bang 幫 initial and the bing 並 initial are represented as [p] or [ph], in that [p] occurs more frequently than [ph]. The pang 滂 initial is represented as [ph] or [p]. The ming 明 initial is represented systematically as [m] in EMSK. There are two

合 開

合

合

開 開

開 開 開 開

開

開

開

開

戈 麻

模

灰

泰 皆

佳 夬 祭 齊

支

脂

豪

肴

pi 碑羆 pʰi 彼陂 pi 匕妣庇畀比 pʰi 秕

pi 鄙轡悲秘閟

uo

uoi

ɑi ɣ ɛi

ɛ æi jᴇi ei

iᴇ

iᴇ

i

1

1

1 2

2 2 3 4

3

3B 3

ɑu

æu

1

2

ɣ

ɣ

3B

i

ɣ

ɣ

ɣ

ɑ ɣ æ

po 堡褓寶報保 pʰo‍ 褒 pʰo 飽苞包 pʰio 豹‍‍

pʰiəi 蔽 pi 篦蓖螕鎞箅 pʰiəi 嬖閉 pi 卑屄髀臂裨

pʰai 貝 pʌi 拜

pʌi 背輩杯

w

1 2

幫 p pʰa 簸波菠播 pʰa 弝欛靶芭笆疤把 pʰai 壩霸 po 圃i逋補 pʰo 布晡ii

Table 3.1  The group bang 幫 in EMSK

pʰo 砲疱‍‍泡‍脬抛‍‍

pʰi 鈹披 pi 䊧 pʰi 屁

pʰi 批v

pʰai 派

pʰai 沛

po 浦iii普 pʰu 鋪 pu 溥 pʰo 怖 pʌi 坯醅胚配

滂 pʰ pʰa 坡破頗 pʰa 帕𧵅葩舥怕

pʰo 皰‍庖‍匏‍咆‍炮‍跑‍麅‍齙鉋‍

pʰo 暴菢袍抱

pʰai 稗𪐄牌𥱼 pʰai 敗 pʰiəi 弊幣 pi 鼙vi pʰiəi 陛梐 pi 脾婢 pʰi 避 pʰi 被鞁vii髲疲皮 pi 鼻琵毘枇 pʰi 鈚膍 pɨi 圮 pi 備

pʌi 排 pɨi 憊 pʰai 鞴

po 捕步𩊶捗 pu 簿埠 pʰo 哺酺蒲葡iv蒱 pʌi 焙培陪倍徘 pʰai 佩悖

並 b pʰa 蔢婆櫇 pʰa 琶耙爬杷罷

mo 貌‍蝥‍茅‍

mi 眉麋湄美媚 mʌi 魅 mo 帽毛瑁旄耄冒媢

mi 糜𪎭靡縻 mʌi 寐

mi 獼麛瀰弭彌

miəi 袂 mi 米迷㴹

mʌi 埋霾 mai 邁 mai 賣 mʌi 買蕒

mʌi 每妹昧苺黴梅媒煤莓酶

mo 姥暮模氁慕謨 mio 墓 ma 媽

明 m ma 魔磨蘑摩 ma 瑪螞馬罵麻蟆

開

開

開 開 開 開 合

開 開 開

開

開 開

開 合

合

宵

侯

尤 幽 鹽 侵 桓

末 山 黠 刪

仙

薛 先

屑 魂

沒

iᴇu u

ɨu iu ɣ iᴇm ɣ im w ɑn

3B 1

3 3 3B 3B 1

ɣ

ɑt ɛn ɣ ɛt ɣ æn

iᴇn ɣ iᴇn iᴇt en

et uon

uot

3 3B 3 4

4 1

1

w

2 2 2

ɣ

iᴇu

3

pon 畚本 pun 錛奔

pal 鉢 pan 扮 pʰal 八 pan 螌頒 pʰan 板 pʰiən 鞭編 piən 變 piəl 鼈虌別 piən 邊藊遍徧 pʰiən 蝙

pʰiəm 貶 pʰum 稟 pan 般半

pʰio ‍表‍臕‍鑣‍𦔩

pʰio 標

pun 噴

pʰiən 片

pʰiən 騙篇偏

pan 攀

pʰum 品 pan 冸 pʰan 判 pən 潘 pal 醱

pu 覆

pu 剖

pʰio 縹剽飄漂

pʌl 鵓脖𩓐荸

pun 盆

piən 玭 pʰiən 艑xii

pʰiən 便楩 piən 辯弁辨

pal 鈸拨 pʰan 瓣辦

pan 伴x盤xi瘢蟠叛磻

pu 浮

pu 部抔裒

pʰio‍ 瓢‍薸‍鳔

miəl 篾蠛 man 暪 mun 門捫 min 悶 mol 沒

miən 面綿 miən 娩冕免勉 miəl 滅搣 miən 眠麵眄

man 蠻慢

mal 抹沫䴲末

(Continued )

man 漫幔蔓鰻鏝饅鞔謾滿曼

mio 緲眇妙渺 tsʰio 杪viii mio 廟‍苗‍描‍貓 mo 某牡拇母畝姆 mu 貿楙 茂瞀袤 mo 矛眸鍪麰謀 liu 繆ix

昔

開 開 合 開

陌 清

ɣ

2 3

開 開

麥 庚

ɛŋ

ɣ

開

耕

2

2 2

開 開

3 3 3 3

1

3

in it ɣ it ɑŋ ɑk ɣ ɔŋ ɣ ɔk əŋ ək ɨŋ ɨk ɣ æŋ ɣ æk

3B 3 3B 1 1 2

piŋ 聘 piək 僻癖

piək 碧 piəŋ 餅并

piək 璧襞辟

ɣ

ɣ

jæk iᴇŋ w iᴇŋ iᴇk

piək 堛xvi pʰʌiŋ 烹 pʌik 魄珀拍

paŋ 滂磅 pak 粕膊濼 pʰaŋ 肨xiii pak 璞朴

pʰil 匹

piək 薜 piəŋ 柄兵丙炳秉

pʌik 伯百栢 pak 迫 pɨŋ 繃xviii䙀xix piəŋ 迸

pʰil 觱必畢 pʰil 筆 paŋ 蒡榜謗 pak 餺愽煿牔博搏 paŋ 邦 pak 駮 pɨŋ 崩 pɨk 北 piŋ 冰 piək 逼

pin 殯鬢賓擯

ɛk jæŋ

ɣ

in

3

開 開 開

開 開 開

開

唐 鐸 江 覺 登 德 蒸 職 庚 陌

質

真

Table 3.1 (Continued)

piək 闢

piəŋ 病評 pʰiəŋ 坪枰平

pʌik 帛舶白

paŋ 傍螃膀 pak 箔泊礡薄 paŋ 棒蚌 pak 雹骲鰒𩪛 pɨŋ 朋堋棚xv pok 蔔 pɨk踣 piŋ 憑 pʰip 煏xvii pʰiak 愎

piŋ 蘋嬪牝 pin 獱㰋嚬頻 pin 貧 pʰil 柲

miəŋ 名

mʌiŋ 萌氓甍 miəŋ 黽 mʌik 脈麥 miəŋ 皿命明鳴 mʌiŋ 盟

mʌiŋ 猛盲虻孟 mʌik 陌貊

mɨk 墨黙

min 旻敏憫愍 mil 蜜謐 mil 密 maŋ 蟒茫忙莽芒邙 mak 幕膜漠莫寞 paŋ 厖xiv mak 邈

min 泯民

開 合

合

錫 東

屋

eŋ

ek uŋ juŋ uk iuk

4

4 1 3 1 3

h

[p ]:34 [p]:31

[p]: 92 [pʰ]:43 h

pok 醭

pok卜 h

piək 霹

piək 壁

6

5

4

3

2

1

Chapter [p]:843 Page 23 Line 4 from the bottom 4 Page 24 [ph]:62 Chapter Chapter 6 Page 9

Chapter 3 Page 4 Line 3 Chapter 4 Page 74

Page 1 pok 僕 Chapter pʰok 4曝

Chapter 3 Page 10

pʰiəŋChapter 萍 3 Page 17 Line 12 piəkChapter 甓 4 Page 31 Line 15 Chapter 3 Page 29 Line 1 from the bottom poŋ 蓬篷 Chapter 4 Page 47

After

mok 目苜牧  睦 [m]:183 [p]:1 [l]:1 [tsh]:1

miək 覓 moŋ 矇蠓蒙懵 moŋ 夢 mok 木沐

miəŋ 茗暝榠瞑冥銘溟

Before

h

Chapter 3 Page 19 i HMJH (YS edition) [po], HMJH (DD edition) [p o], HD [p u], SU, JU [p o] 7 Chapter 4 Page 75 ii HMJH [pho], HD [pu] (popular reading[pho]) iii HMJH (YS edition) [po], HMJH (DD edition) [pho], YH [po], HD [phu], SG, SU, JU8 [pho] Chapter 3 Page 27 Line 10 Chapter 4 Page 44 iv HMJH [pho], HD [pu] (popular reading [pho]) h v HMJH [p i], HD [pʰiəi] (popular reading [pi] Chapter 3 Page 23 Line 2 from the bottom 9 Chapter 4 Page 29 vi HMJH [pi], HD [pʰiəi] (popular reading [pi]) vii HMJH [phi], SU, JU [pi] Chapter 3 Page 2 10 鎞 viii YH [tshio], HD, OP [mio] (popular reading [tshio]) Chapter 4 Page 16 ix YH [liu], HD [mu] (popular reading [liu]) Chapter 3 Page 16 11 x HMJH [pan], SG[phan] Chapter 4 Page 2 xi Although the phonetic notation for this character is [span] in HMJH (YS edition, DD edition), this is probably a mistake, since the [sp] consonant cluster is not seen Chapter 3 Page 25 elsewhere. The phonetic notation for this character is [pan] in DH and YJ. 12 Chapter 4 Page 72 xii HMJH [pʰiən], SU, JU [piən] xiii HMJH [pʰaŋ], HD, SU, JU [paŋ] xiv HMJH, HD, SG [paŋ], OP [maŋ] (standard reading [paŋ]), SU, JU [maŋ] xv HMJH, HD [pɨŋ], OP [phʌiŋ] (standard reading [pɨŋ]), SU, JU [phʌiŋ] xvi HMJH [piək], HD [phik], OP [piək] (standard reading [phik]) xvii HMJH [pʰip], HD [phik], OP [piək] (standard reading [phik]), SU, JU [piək] xviii HMJH, HD [pɨŋ], SU, JU [phʌiŋ] xix HMJH, HD [pɨŋ], SU, JU [phʌiŋ]

Summary

開

青

Location Chapter 3 Page 7 4 Page 43 Line 3 piəŋ Chapter 瓶屏軿並

62  Initial system fanqie 反切 spellings for 杪 in Jiyun 集韻, which are 彌沼 and 楚敎. The reading of 杪 [tsʰio] is consistent with the 楚敎 spelling. The reading of 繆 is [liu] in YH and [mu] (popular reading [liu]) in HD. The reading of 繆 [liu] may derived from the analogy of the lai 来 initial character 镠 [liu]. The reading of 厖 [paŋ] is probably derived from the analogy of its phonetic component pang 尨, which is a variant form of the bing 並 initial character 龐 [paŋ].

3.2 Group fei 非 (labiodentals) Table 3.2  The group fei 非 in EMSK 非f pu 府俯斧賦傅夫 膚跗柎付 pʰo 脯 Chapter 3 Page 11 4 Page 15 3 wjai 廢 合 Chapter pʰiəi 廢 bottom ɨi 12 from 3 24wLine pi the 榧篚沸扉非飛匪 微 Chapter 合 3 Page Chapter 4 Page 43 Line 15 ii pɨl 疿 Chapter 3 Page 20 pʌi 緋iii Chapter 4 Page 20 3 ɨu 尤 開 pu 缶殕否缹富𦉓 虞

13 14 15 16 17 18

凡 乏 元

月 20

22

3 jṷo̝

文

pɨl 不芾弗 paŋ 舫方昉放

3 juŋ 3 juk

pʰuŋ 諷楓風𩘲 pok 幅福腹蝠辐

pʰuŋ 豐 pok 蝮

3 jṷo̝ŋ

poŋ 封

ṷjap

Chapter 3 Page 4 Line 5 from the bottom 3 wiɤn 合 pən 藩蕃 Chapter 4 Page 48 Chapter 5 Page 3 Line 13 & 22 iv

pʰan 販 畈阪v Chapter 3 Page 5 pan 反返 Chapter 4 Page 7 3 3wiɤt 合 Chapter pal3 髮發 Page 23 Line Chapter 4 Page 3 Chapter 5 Page 3 Line 10

3 3jun pun6 Chapter Page 24 Line Chapter 4 Page 35

合

3

東 屋

合 合

合 燭 合 Summary i ii iii iv v

粉糞分饙奮

Chapter 3 Page 12 Line 15 Chapter 4 Page 37 Line 3

物 陽

鍾

奉 v 微 ɱ pu 父腐釜鮒 mu 鵡武舞廡 駙符芙鳧 務霧巫誣無 扶附 po 輔 侮騖蕪 pʰiəi 吠 pi 翡剕腓肥 mi 尾味薇微

jut iaŋ

w

pu 副

彀

3 3ṷjam 合 Chapter Page 29 Line 8 Chapter 4 Page 37 Line 10 from the bottom

合

敷 fʰ pu 俘郛麩稃 莩筟敷赴 mu 撫i pʰiəi 肺 pi 妃費

pu 婦負阜蝜 蜉芣𣑿浮 pəm 泛 pəm 帆凡犯範 pəp 法 pʰip 乏 pan 飯 pən 旙 pən 燔蘩 轓翻 礬𢶃繁煩 𣂐𣂐 pəl 罰筏瞂 昁伐 pun 紛忿 pun 坌𢅯墳 𢻨𢻨 枌濆豶鼢 焚憤 𧞣𧞣 pɨl 拂 pul 佛pɨl 咈 paŋ 紡芳訪 paŋ 房魴防

Chapter 3 Page 6 Chapter 4 Page 31 Line 5

19

21

合

3 jṷo̝k [p]:49 [ph]:9

man 晚萬挽 mal 韈 mɨn 吻問璺聞 文蚊紋 mun 紊 mɨl 物 maŋ 網輞鋩望 亡罔忘妄

poŋ 鳳 pok 服鵩復 馥伏 poŋ 捧蜂鋒 poŋ 縫逢奉 烽丰峯 pok 幞 [p]:29 [ph]:2 [p]:63 [m]:1 [ph]:2

HMJH, YH, SE, HD [mu], SU, JU [pu] HMJH [pɨl], Modern SK [pi] HMJH [pʌi], HD, SU, JU [pi] HMJH [pʰan], YH [pan] HMJH [pʰan], HD [pan] (popular reading [phan])

[m]:37

Initial system 63 The fei 非 [f], fu 敷 [fʰ] and feng 奉 [v] initials, which are derived from the bang 幫 [p], pang 滂[pʰ] and bing 並 [b] initials, respectively, are represented as [p] or [ph] in EMSK. The reading of 撫 [mu] results from the analogy of its phonetic component 無 [mu]. The wei 微 [ɱ] initial, which is derived from the ming 明 [m] initial, is represented as [m] in EMSK.

3.3 Group jian 見 (velars) The jian 見 initial is generally represented as [k] in EMSK. The [h] reading in the following characters can be explained by analogy: 亥 (匣 [ɦ] initial) > 荄該 痎 [hʌi], 會 (匣 [ɦ] initial) > 儈膾檜 [hoi], 解 (匣 [ɦ] initial) >廨懈 [hʌi], 合 (匣 [ɦ] initial) > 鴿閤蛤 [hap], 咸 (匣 [ɦ] initial) > 緘 [ham], 狹峽 (匣 [ɦ] initial) > 莢頰筴鋏蛱裌 [hiəp], 汗 (匣 [ɦ] initial) > 旰趕[han], 鎋轄 (匣 [ɦ]) > 割 [hal], 還環 (匣 [ɦ]) > 鰥 [hoan], 現 (匣 [ɦ] initial) > 筧 [hiən], 齕麧 (匣 [ɦ] initial) > 訖 [hɨl], 戄彏䂄 (曉 [h] initial) > 钁 [hoak], 巷 (匣 [ɦ] initial) > 港 [haŋ], 恒 (匣 [ɦ] initial) > 亘 [hʌŋ], 刑 (匣 [ɦ] initial) >荊 [hiəŋ] (Pak 1971: 31; Kōno 1979: 359; Lee 1990: 131–132). There are two fanqie spellings for 涓 in Jiyun, which are guixuan 圭懸 and yingjuan 縈絹. The [iən] reading of 涓 in EMSK is consistent with the fanqie yingjuan 縈絹. The reading of 訐鴶 [al], according to Lee (1997: 117), is a result of the lenition of [k]. The xi 溪 initial is generally represented as [k] in EMSK. The [kh] reading is only seen in 快 [khoai]. The reading of 擴 [hoak] is consistent with the fanqie 忽郭 in Jiyun. The [h] reading of the xi 溪 initial in the following characters can be explained by analogy: 亥 (匣 [ɦ] initial) > 咳 [hʌi], 灰 (曉 [h] initial) > 恢[hoi], 含 (匣 [ɦ] initial) > 龕 [ham], 盍 (匣 [ɦ] initial) > 榼 [hap], 歆(曉 [h] initial) > 欠欽 [hɨm], 狹峽 (匣 [ɦ] initial) > 篋愜 [hiəp], 活 (匣 [ɦ] initial) > 蛞 闊 [hoal], 絜 (匣 [ɦ] initial) > 挈 [hiəl], 恄 (曉 [h] initial) > 詰 [hil], 杭 (匣 [ɦ] initial) > 伉抗 [haŋ], 穫鑊 (匣 [ɦ] initial) > 確 [hoak], 鵠 (匣 [ɦ] initial) > 酷 [hok] (Pak 1971: 38; Kōno 1979: 356–357; Lee 1990: 131–132). The reading of 泣 [ɨp], which is expected to be [kɨp], results from the lenition of the initial (Lee 1997: 117). The qun 群 initial is represented as [k], and the yi 疑 initial is represented as a zero-initial in EMSK. The exceptional readings that result from analogy are shown as follows: 臯 (見 [k] initial) > 翱 [ko], 險 (曉 [h] initial) > 驗 [həm], 亂 ( 來 [l] initial) > 薍 [lan], 檗薜 (幫 [p] initial) > 蘖 [piək], 齕麧 (匣 [ɦ] initial) > 屹 [hɨl], 謔 (曉 [h] initial) > 瘧虐 [hak], 更 (見 [k] initial) > 硬 [kiəŋ].

3.4 Group ying 影 (gutturals) The ying 影 initial is generally represented as a zero-initial in EMSK. In the following characters, the ying initial is represented as [k]: 嫗嘔漚甌鷗謳蓲. As pointed out by Kōno (1979: 360) and Pak (1971: 50), the readings of these characters result from analogy of the phonetic component 區 [ku]. Other readings that result from analogy include 蠻 [man] > 彎 [man] and 鑊 (匣 [ɦ] initial) > 蠖雘 [hoak] (Pak 1971: 50). The reading of 杳 [mio] may derived from the analogy of 渺 [mio], a character that shares a similar meaning with 杳 “obscure” (Lee 1997:

wɣ

ɛ

2

合

ɣ

ɛ

ɛi

wɣ

2 2

合 開

佳

ɛi

ɣ

2

開

皆

ɑi

1

開

泰

uoi

1

合

灰

əi

1

開

咍

jṷo̝

3

合

虞

jɤ

3

合

模

麻

魚

ɑ w ɑ w jɑ ɣ æ wɣ æ uo

1 1 3 2 2 1

開 合 合 開 合 合

歌 戈

kai 蓋丐 hoi 儈膾檜 kʌi 稭皆 kai 疥介芥㠹 kiəi 界階犗誡戒 hʌi 痎偕 koi 乖怪 kai 街 ka 佳 hʌi 廨懈解 koai 拐枴掛

kai 概漑槩 hʌi 荄該 kʌi 改

ka 嫁架稼駕斝椵假家袈茄枷痂笳葭耞嘉加 koa 寡苽蝸瓜 ko 股賈瞽罟鼓蛊牯羖顧雇錮孤沽辜姑箍蛄 古捩菇故固蠱痼 kə 鋸車居踞據裾擧 kəi 筥 ku 句駒𣂐矩拘俱

見 k ka 歌柯哥 koa 菓戈堝腡鍋果裹過

Table 3.3 The group jian 見 in EMSK

kʌi 揩

koi 傀塊魁 hoi 恢

kʌi 開 kai 鎧慨 hʌi 咳

ku 軀驅區

kə 袪去

koa 銙胯誇 ko 苦庫袴枯

溪 kh ka 珂可軻 koa 課騾科蝌窠

kə 距炬苣𦼫渠蕖 𪆂拒巨遽 ku 衢癯臞蠷具懼窶

ka 瘸

群 g

koi 聵i ai 涯崖

ai 艾 oi 外 ai 騃

ui 桅 ɨi磑

u 愚虞隅寓遇 o娛 ai 礙

a 砑牙芽衙雅迓訝 oa 瓦𡧗‍‍ o 五伍寤捂䎸吾蜈鼯杇誤 悟忤迕 io梧 ə 圄御語馭魚漁

疑 ŋ  a 我餓鵞蛾 oa 臥訛

開

開

蕭

侯

豪 肴 宵

開 開 合 開 開 開

之 微

合

開

wɣ

ɨ ɨi w ɨi ɑu ɣ æu iᴇu ɣ iᴇu eu

u

3 3 3 1 2 3 3B 4

1

i

kio 嬌驕矯 kio 徼叫皎 hio 梟 ku 垢笱狗構購覯彀雊溝冓篝鉤韝耈苟

ku 晷 kui 龜 koi 愧 kuəi 軌 kɨi 箕基己紀記 hɨi 姬 kɨi 蟣機譏鞿幾饑既 kui 鬼貴歸 ko 槀誥告羔膏橰櫜篙餻皐高藁 kio 鉸狡酵校教窖覺咬交膠郊攪

æi iᴇi jai ei kiəi 薊髻笄雞計繫繼 w ei kiəi 桂 kiu 閨圭 iᴇ ki 枳 ɣ iᴇ kɨi 徛覉寄 ki 庋 w iᴇ kiu 𨾚規 wɣ iᴇ kuəi 攱𪀗詭 i ɣ i kɨi 肌冀飢覬 kuəi 机 w i kiəi 季

ɣ

wɣ

3B

3 3B 3 3B 3

合

脂

3 3B

開

支

2 3B 3 4 4

合 開 開 開 合

夬 祭 廢 齊

ku 口釦簆鷇寇 摳 ko 叩

ko 槁拷考𥬯犒尻 kio 磽骹巧 kio 蹺 kio 鞽 kiu 竅

kɨi 起欺 kɨi 氣iii 豈

kiu 窺跬 hiu 虧 ki 棄 kɨi 器

ki 企蚑 kɨi 欹綺踦

kiəi 稽啓契鸂溪

kʰoai 快ii kəi 憩揭

彄

ɨi 劓

ui 危僞

kio 轎橋蕎

u 藕腢偶

io 僥

(Continued )

kɨi 棋旗麒錤期忌 ki 其 ɨi 疑擬 kɨi 璣畿祈 ɨi 毅 ui 魏 oi 巍 o 鏊螯熬厫獒傲 ko 翱iv

kuəi 跪 ki 鬐耆 kɨi 跽 kiu 葵揆 kiəi 悸 kuəi 蕢櫃饋 kiu 逵

ki 歧 kɨi 錡騎奇 ki 技妓芰 ɨi 蟻義儀宜議

iəi 藝 iəi 乂 ai 刈 iəi 猊霓鯢詣

3B

3B 1

開

開 開

緝 寒

曷

kɨp 給級汲急 kan 稈笴幹干肝竿奸漧 han 旰趕 kal 葛 hal 割

ip ɑn

4

開

ɑt

ɣ

kəm 劍 kəp 劫 kiəm 蒹縑兼 hiəp 莢頰筴鋏蛱 kɨm 錦今襟金禁

jam jap em ep ɣ im

3

3B

開

開

2

kiəm鹻 kam 減 ham 緘 ɣ ɛp hiəp 裌夾 ɣ æm kam 鑑監 ɣ æp kap 甲胛 ɣ iᴇm kəm 瞼檢

嚴 業 添 帖 侵

洽 銜 狎 鹽

2

開

hap 鴿閤蛤 kam 甘柑泔

əp ɑm ɑp ɣ ɛm

1

kiu 糾 kam 紺感敢

iu əm

3 1

ku 九韭廄鳩鬮久救究

ɨu

3

開

開 開

幽 覃

合 談 盍 咸

開

尤

Table 3.3 (Continued)

kal 渴

hɨm 欠 kəp 怯 kiəm 歉謙慊 hiəp 篋愜 kɨm 衾 hɨm 欽 ɨp 泣vi kan 看刊

kiəp 掐 kam 嵌

kam 瞰坩 hap 榼

kam 坎墈堪 ham 龕

ku 糗丘蚯

kɨp 及

kɨm 妗禽琴檎噙

kam 芡 kəm 儉黔 kiəm 鉗 kɨm 鈐

ku 舅臼柩裘仇毬梂 求舊咎

al 枿

an 岸

ɨm 吟

əm 釅嚴儼 əp 業

həm 驗v

am 巌

u牛

鎋 仙

開

月

屑

先

開 合

元

合 開 合 開

開

薛

合

合 合 開

æn

2

開

2

開 合 開

iᴇn iᴇn

3 3B

4

4

w

iɤt en w en et

iɤt

3

kal 羯 al 訐x kuəl 蹶蕨鱖瘚 kiən 繭襺趼毽見肩鵑堅 hiən筧 kiən 畎蠲 iən涓xi kiəl 鍥𢎹結潔 kil 桔䓀

iᴇt hiəl 孑viii iᴇt iɤn kən 犍鞬建 w iɤn kuəl 厥

ɣ

koan 管輨盥館鸛罐觀官棺冠莞涫觀貫灌 koal 括筈 kan 簡㶕襇間艱揀 hoan 鰥 kal 圿 al 鴶vii kan 澗諫 koan 菅 koan 關慣 koal 刮 kiən 甄 kən謇瀽 kiən 囝 kiən 絹 kuən 眷卷

3A 3B 3

wɣ

w

wɣ

wɣ

æn æt iᴇn ɣ iᴇn

ɣ

2 3 3B

2

ɑn ɑt ɣ ɛn wɣ ɛn ɣ ɛt

w

w

1

合

刪

黠

桓 末 山

kuəl 闕 kiən 牽 kiən 犬 hiəl 挈

kuən 勸券

kiən 譴遣 kən 攐愆

koan 寬款欵 hoal 蛞闊 kan 慳

kuəl 橛

kal 碣竭

kəl 傑 kən 鍵健 kuən 圈

koan 顴 kuən 拳權倦

kən 乾虔件

iəl 霓 hiəl 齧xii

uəl 月刖 iən 硯研妍

(Continued )

əl 糵孽闑 piək 蘖ix ən 言 uən 原黿芫蚖元願源 oan 阮

ən 彦

lan 薍 oan 頑

an 贗鴈顏

an 眼

oan 玩

kaŋ 𤭛崗綱剛岡

koaŋ 光胱廣 kak 閣胳各 koak 彍槨郭

w ɑŋ ɑk w ɑk

3

1

1

合

開

陽

鐸

合 開 合

開 合

3

kiul 橘 kɨn 謹槿靳斤筋 kəl 吃 hɨl 訖xiv kun 君軍皸莙

it ɨn ɨt jun jut ɑŋ

iaŋ kaŋ 襁薑疆姜韁礓𨏃僵 iaŋ koaŋ 誑

w

w

3 3

術 殷 迄 文 物 唐

合 開

開 合

質 諄

uot in ɣ in il w in

1 3A 3B 3A 3

合 開

沒 真

kon 棍滾昆蜫袞鵾 kun 裩 kol 鶻骨榾 kin 緊 kən 巾 kil 吉 kiun 麕鈞均

uon

1

合

魂

kɨn 根跟

ən

kiəl 決 hiul譎

1

開

痕

et

w

合

屑

4

Table 3.3 (Continued)

kul 𨧱屈 kaŋ 炕穅康慷 haŋ 伉抗 koaŋ 纊壙曠 kak 恪 koak 籗鞹 hoak 廓擴xvi kaŋ 羌蜣 koaŋ 眶筐𨴑匡恇

kəl 乞肐虼

hil 詰xiii kiun 囷

kul 窟

kɨn 齦 kʌn 墾懇 kon 閫坤壼困悃

kiəl 缺

kaŋ 糨強 koaŋ 誆狂

kun 郡裙群 kul 掘

kɨn 近芹勤

kiun 菌 kun 窘

kɨn 僅

aŋ 仰

ak 腭鍔蕚鶚諤愕

aŋ 枊

ɨn 垽齗 hɨl 屹xv

ɨn 銀誾

ol 兀

開 開

開 合

清

w

ɣ

ɣ

iᴇŋ kiəŋ 勁 iᴇŋ kiəŋ 頸

wɣ

ɛk

æŋ æk wɣ æk ɣ ɛŋ ɣ ɛk

ɣ

wɣ

3 3

2 2 2 2

合 開 合 開 開

kuk 國 kɨŋ 矜 kɨk 棘襋 kiəŋ 梗綆莄粳鶊鯁 kʌiŋ 羹更秔 koiŋ 觥 kak 骼 kiək 格 koik 虢 kiəŋ 耕耿 kiək 膈𥴩隔 hiək 革 koik 馘 kuk 膕蟈 kiəŋ 景境敬鏡京蟼驚警竟 hiəŋ 荊 kɨk 戟

jæŋ jæk

2

開

ək ɨŋ ɨk ɣ æŋ

w

3 3

3

合 開

合

1

hʌŋ 亘 koiŋ 肱

əŋ əŋ ək

1

w

iak kak 腳 iak hoak 钁 ɣ ɔŋ kaŋ 講洚绛江扛矼豇 haŋ 港 ɣ ɔk kak 角桷覺

w

3 3 2

開 合 開

開 合 開

庚 陌

耕 麥

陌

蒸 職 庚

德

登

覺

江

藥

kiəŋ 卿慶 kiə 綌 kɨk 隙 kiəŋ 輕 kiəŋ 檾傾頃

kiəŋ 瓊

iəŋ 迎 iək 逆

kiəŋ 硬xix

kʌiŋ 鏗

ɨŋ 凝

ʌik 額

kiəŋ 鯨檠黥𢐧擎競 kɨk 屐劇

kɨk 極

ak 嶽樂

hak 瘧xvii虐xviii

kʌik 客

kʌiŋ 坑

kʌk 刻 kɨk 克

kak 慤 hoak 確 kɨŋ 肯

kaŋ 腔

kak 却

(Continued )

合

錫 東

屋

i ii iii iv v vi vii viii ix x xi xii xiii xiv xv xvi xvii xviii xix

4 1 3 1 3 1 3 3

4

eŋ eŋ ek uŋ juŋ uk juk ok jṷo̝ŋ jṷo̝k

w

[k]:425 [h]:36 [ø]:3

kiəŋ 徑經涇 kiəŋ 扄 kiək 墼激擊 koŋ 貢功公工蚣工攻 kuŋ 宮弓躬 kok 谷穀轂 kuk 鞠菊掬鵴 kok 梏 koŋ 拱恭筇供鞏 kik 喫 koŋ 孔鞚𧌆控空 kuŋ 穹 kok 哭 kuk 麴 hok 酷 koŋ 恐 kok 曲蛐𦚼 [k] 164 [kh]:1 [h]:19 [ø]: 1

kiəŋ 磬罄

koŋ 蛬共 kuk 局跼 [k]:114

kuŋ 窮

HMJH also lists [huəi], HD [hoi], SU, JU [oi] YH, IG [khoai] The notation in HMJH (YS edition) is [skɨi], but this is probably a mistake, since the [sk] consonant cluster is not seen elsewhere. HMJH, YH, SG [ko], HD, OP [o] (popular reading [ko]), SU, JU [o] YH [həm], HD [iəm] (popular reading [həm]), OP [əm] (popular reading [həm]) HMJH, YH, SE, IG, SG [ɨp], HD, OP [kɨp] (popular reading [ɨp]), SU, JU [kɨp] HMJH [al], Modern SK [al] HMJH, SG [hiəl], HD [kiəl], OP [kiəl] (popular reading [hiəl]), SU, JU [kəl] HMJH [piək], HD, SU, JU [əl] HMJH, YH [al], HD, OP [kal] (popular reading [al]), SU, JU [kal] YH [iən], HD, OP [kiən] (popular reading [iən]) HMJH, YH [hiəl], SU, JU [əl], HD [iəl] (popular reading [siəl]), OP [əl] (popular reading [siəl]) YH [hil], HD, OP (popular reading [hil]) YH, SE, HD [hɨl] YH [hɨl], HD, OP [ɨl] (popular reading [hɨl]) YH [hoak], HD, OP [koak] (popular reading [hoak]) HMJH, HD [hak], OP [iak] (standard reading [hak]), SU, JU [iak] YH, SE, SG, SH, HD [hak], SU, JU [iak], OP [iak] (standard reading [hak]) YH [kiəŋ], HD, OP [iəŋ] (popular reading[kiəŋ])

沃 合 鍾 合 燭 合 Summary

開 合 開 合

青

Table 3.3 (Continued)

ok 獄玉 [ø]:136 [h]:5 [k]:3 [l]:1 [p]:1

ʌik 鷁

開

合

開 合 開 合 開 合 開 合 合 合 開 合

咍

灰

泰

祭 廢 齊

夬

佳

皆

合 合

開 合 合 開 合 合

魚 虞

模

麻

歌 戈

ɑ w ɑ w jɑ ɣ æ wɣ æ uo

jɤ jṷo̝

əi

uoi

ɑi w ɑi ɣ ɛi wɣ ɛi ɣ ɛ wɣ ɛ ɣ æi wɣ æi w jᴇi w jai ei w ei

1 1 3 2 2 1

3 3

1

1

1 1 2 2 2 2 2 2 3 3 4 4

iəi 濊穢 iəi 曀翳

ai 餲

ai 矮 ʌik 隘

ai 噫挨

ai 藹

ə 飫於 ku 嫗 u紆o迂 ʌi 愛哀 ai 埃 oi 椳偎煨

a 瘂鴉 oa 蛙窪搲 o 烏洿𦶀汙

影ʔ a阿 oa 渦倭窩萵

Table 3.4  The group ying 影 in EMSK

huəi 喙 hiəi 醯

koi 歪ii

hoi 賄誨灰晦悔

hʌi 海醢

hə 虛噓許 hu 酗煦

曉 h ha hoa 火貨夥 hoa 靴 ha 罅 hia 嚇 hoa 化樺花 ho 琥戽虎呼

kiəi 系蹊 hʌi 嵇 hiəi 慧惠 kiu 畦 hiu 携

hoa 話

hoi 茴廻匯回徊 koi 槐i kuəi 潰 hai 害 hoi 繪會 hʌi 骸諧駭 hiəi 薤 hoi 懷 koi 壞 hʌi 蟹邂 hiəi 鞋 hoa 畫

hʌi 孩頦

ha 下廈夏遐瑕蝦霞鰕ka暇 hoa 華鏵 koa 踝 ho 户帍簄枑狐瑚壺猢湖弧糊箶衚鬍鶘瓠蝴護互 胡乎冱

匣 ɦ ha 河荷魺何苛賀 hoa 禾禍和

ui 衞

(Continued )

u 羽雨宇芋盂栩禹

云 ɦ>j

開

支

脂

開 開 合

開

開

開

開

開

開

開

之 微

豪

肴

宵

蕭

侯

尤

幽

合

開 合

iᴇ iᴇ wɣ iᴇ

i ɨi w i

i

ɨ ɨi

ɑu

æu

ɣ

iᴇu iᴇu eu

u

ɨu

iu

3 3B 3

3 3 3

1

2

3 3B 4

1

3

3

ɣ

wɨ

w

ɣ

3B 3 3B

Table 3.4 (Continued)

iu 幼幽

‍ 邀 腰 io 夭‍妖 mio 杳vi lio 窈vii ku 蓲嘔 漚甌鷗 謳 u 優耰麀憂

io ‍坳‍‍ lio 拗

o 襖‍媼‍爊奧

ɨi 意薏醫 ɨi 扆衣依 ui 威蝛慰 oi 畏

i 蛜伊 ɨi 懿 əi 饐

ɨi 椅倚 ɨi 恚 ui 餵萎委

hiu 鵂休 hu 朽嗅

hu 吼

hio 曉

ho 好‍薅‍蒿 mo 耗iv ho‍ 髇‍‍ hio ‍孝‍哮

hɨi 蟢禧喜 hɨi 豨希稀 hui 暉楎輝諱揮 huəi 卉虺

hɨi 戲曦 hiu 觽 huəi 毀燬 hui 麾

hu 後后鍭候堠侯猴糇喉帿逅厚

hio‍ 肴‍‍效

ho 昊‍號‍豪‍壕‍嘷浩毫‍灝皓

u 右友祐疣尤又 iu 囿有宥

hio‍ 鴞‍

ui 葦蝟緯胃韋幃闈 煒謂偉違圍渭 hui 彙iii

iu 帷鮪 ui 位

ui 為

開 開 開

業 添 帖 侵

末

寒 曷 桓

緝

開

3 3B 1

1

開

合

4 4 3B

iᴇp jam

3B 3

葉 嚴

w

ɑt

ip ɣ ip ɑn ɑt w ɑn

jap em ep ɣ im

ɣ

3 3B

開

ɨm 飲窨音陰蔭 am 瘖 ɨp 揖挹 ɨp 邑浥 an 案鞍安按 al 遏閼 uən 盌 oan 腕豌帵 al 斡

əm 淹

ap 鴨壓 iəm 黶檿𧞣厭 əm 閹醃晻奄掩 am 俺

am 黯

ap 罨

əp ɑm ɑp ɣ ɛm ɣ ɛp ɣ æm ɣ æp iᴇm ɣ iᴇm

am 暗庵䳺諳

əm

1 1 2 2 2

1

開 開 開 開 開

開

合 談 盍 咸 洽 銜 狎 鹽

覃

hal 豁

hoan 獾歡喚煥渙

hɨp 吸 han 罕漢

hɨm 歆

hɨm 杴 həm 蘝 həp 脇脅

həm 險

ham 憨

hoal 活

han 旱悍銲鼾汗翰𤿧寒韓捍 kal 毼 hoan 換紈芄萑桓丸 oan 完xi緩xii澣xiii

hiəm 嫌 hiəp 協挾俠

ham 撼頷含函涵 kam 憾 hap 䤴viii hap 合 kam 酣 hap 闔 ham 餡陷鹹咸 hiəp 狹 hɨp 洽 ham 艦檻銜 kap 匣 ap 狎ix

(Continued )

niəp 饁x燁

iəm 炎

痕 魂 沒 真

屑

先

月

元

鎋 仙

刪

黠

山

ɛn ɛt wɣ ɛt ɣ æn wɣ æn ɣ æt ɣ iᴇn w iᴇn iɤn w iɤn

2 2 2 2

合 開 合 開 合 開 開 合 開 合

1 1 1 3 3B

4

合 開

合 開 合 合 開

4

開 合 開

3

2 3B 3 3

ɛn

w

et ən uon uot in ɣ in

w

en et

iɤt w iɤt en

ɣ

wɣ

ɣ

2

開

Table 3.4 (Continued)

in 印姻茵因

ɨn 恩 on 搵瘟溫穩

iəi 蠮

ən 堰鰋 oan 畹婉 uən 苑踠 鴛冤怨 al 謁 əl 噦 iən 嚥鷰宴煙胭 咽淵

ən 焉

an 晏 man 彎xiv

hɨn 釁

hon 昏婚閽惛 hol 笏忽惚

hiəl 血

hiən 蜆顯

hən 獻軒憲 huən 楦萱 暄貆喧 hun 塤 həl 蠍歇

hal 瞎

hɨn 痕狠 hʌn 恨 hon 混渾圂魂餛 kol 𣝗 xv

hiən 玄衒懸眩 hil 翓 hɨl 齕麧 hiəl 穴

hiən 峴莧賢弦舷縣絃

kan 癇 han閑 hʌn 限 hoan 幻 hoal 猾 hil 黠 hoal 滑 han 骭 hoan 宦豢槵環寰鬟還患 hal 轄

uəl 鉞越 oal 曰

uən 遠猿園垣轅榬援

uən 院衏員圓

開 合

殷 文

蒸 職

德

登

覺

江

藥

陽

鐸

開 合 合 開 開

開 合 開 合 開 合 開

開 合

開

質

物 唐

合

諄

3

1

1

2

3

1

əŋ w əŋ w ək ək ɨŋ ɨk

ɣ

ɔk

ɑk w ɑk iaŋ w iaŋ iak w iak ɣ ɔŋ

ɨŋ 鷹膺應 ək 億臆憶抑

ak 幄渥握

ak 堊惡 hoak 蠖xvii雘xviii aŋ 秧鴦胦央殃怏 oaŋ 枉 iak 約

ul 蔚熨鬱 aŋ 盎 oaŋ 汪

jut ɑŋ w ɑŋ

1

il 一壹 ɨl 鳦 ɨn 癮殷隱 on 緼醖蘊

it ɣ it ɨn jun

3 3B 3 3

in

w

3

hɨk 黑 hɨŋ 興

huŋ 薨

hiaŋ 𢴦 haŋ 肛 hʌk 鷽學 kak 确 hʌŋ 恒 hoŋ 弘 hok 惑 hʌik 劾

haŋ 項巷缸降xix

haŋ 𥮕行航吭䘕𦐄 hoaŋ 謊荒 hoaŋ 幌黃皇隍凰篁蝗潢蟥璜惶煌 haŋ 肓 hak 壑 hak 鶴貉 hoak 涸 hoak 癨臛 hoak 穫鑊 hiaŋ 饗鄉香響向享 hoaŋ 怳貺

hɨn 脪昕欣 hun 訓葷纁勳薰

iak 籰

oaŋ 王往

(Continued )

un 暈餫耘雲云韻 紜運

kiun 筠xvi un 隕

juŋ uk

juk ok

3 1

3 1

合

合

屋

沃

錫 東

4 1

iᴇŋ iᴇk eŋ w eŋ ek uŋ

3 3 4

合 開 開 合 開 合

清 昔 青

庚

麥

ɛŋ hoiŋ 泓xx ɛk ʌik 軛厄 wɣ ɛk ɣ jæŋ iəŋ 影映英 wɣ jæŋ

合 開 合 開 合

w

ɣ

wɣ

wɣ

uk 栯燠 ok 沃鋈

ok 屋

oŋ 塕齆甕翁螉 䩺甕

iəŋ 癭瓔纓嬰縈 ik 益

hoŋ 𦕠

hiəŋ馨

hiəŋ 兄

hiək 赫

2 2 2 3

ʌiŋ 鸚鶯櫻 iəŋ 甖

陌 耕

æŋ æk ɣ ɛŋ

ɣ

2 2

合 開 開

iək 閾 hiəŋ 亨

ɣ

ɨk æŋ

w

3 2

合 開

職 庚

Table 3.4 (Continued)

kok 鵠xxii

kok 槲 kok 斛縠

hiəŋ 脛刑形型硎 hiəŋ 螢迥炯 hiək 覡 hoŋ 汞閧紅 洪虹鴻烘篊葒

kʌiŋ 莖 hʌiŋ 幸倖 koiŋ 紘 koiŋ 宏 hʌik 覈核 kiək 翮 hoik 獲畫

hʌiŋ 杏行胻桁 hiəŋ 衡 hoiŋ 橫

uŋ 雄熊

iəŋ 泳蝾榮永嶸咏詠 hiəŋ 瑩xxi

iək 域

jṷo̝k

3

燭 合 Summary

i ii iii iv v vi vii viii ix x xi xii xiii xiv xv xvi xvii xviii xix xx xxi xxii

[ø]:207 [k]:7 [h]:3 [m]:2 [l]:2

oŋ 壅癰饔𪄉雍 壅擁 uk 旭勖 [h]:132 [ø]:3 [k]:1 [m]:1

hiuŋ 胸凶

[h]: 228 [k]:23 [ø]:4

YH, SU, JU [hoi], HMJH, CJM [koi], OP [hoi] (popular reading [koi]) HMJH (YS edition) [koi], HMJH (DD edition) [oai] HMJH, IG, HD [hui], OP [ui] (standard reading [hui]), SU, JU [ui] YH [mo], OP [ho] (popular reading [mo]), Modern SK[mo] HMJH, YH, SE, SG [hio] CJM [mio], HD, OP [io] (popular reading [mio]) In addition to [lio], YH also lists the reading [mio]. HMJH (YS edition and DD edition) [hap], HMJH (GMH edition) [ham] YH, SE, HD [ap], OP [hap] (popular reading [ap]) HMJH [niəp], SG [iəp] YH, SG [oan] YH, IG [oan] In addition to [oan], YH also lists the notation [han]. SG [han], HD [oan] (popular reading [han]), OP [hoan] (standard reading [oan]) HMJH [man], HD, OP [oan] (popular reading [man]), SU, JU [oan] HMJH (YS edition) [kol], HMJH (DD edition) [hol] HMJH [kiun], HD, OP [iun] (popular reading [kiun]), SU, JU [iun] HMJH, SE [hoak] HMJH, YH [hoak] In addition to [haŋ], YH also has the notation 降 [kaŋ]. YH [hoiŋ], HD, OP [hoiŋ] (popular reading [hoŋ]) SG [iəŋ], YH [hiəŋ], HD [iŋ] HMJH, YH, SG, JY, HD [kok], SU, JU [hok]

jṷo̝ŋ

3

合

鍾

[ø]:72 [h]:3 [k]:1 [n]:2

78  Initial system 120). The reading of 拗窈 is [lio] in YH. The [l] initial of 拗窈 [lio] may derived from the coda [l] of the preceding syllable in the notation in YH (Lee 1997: 120). The initial of 泓 [hoiŋ] may arise from the initial of its phonetic radical 弘, the initial of which is 匣 [ɦ]. The xiao 曉 initial is generally represented as [h] in EMSK. The readings that result from analogy include 毛 (明 [m] initial) > 耗 [mo] and 域 (云 initial) > 閾 [iək]. The reading of 旭勖 [uk] is a reflection of the lenition of the initial [h]. The [h] initial, which shares similar place of articulation with [u], was dropped due to the effect of dissimilation. The xia 匣 initial is generally represented as [h], and occasionally as [k] in EMSK. The reading of 完緩澣 [oan] drops the initial [h] due to the lenition of the initial. There are two fanqie spellings for 壞 in Guangyun 廣韻, namely guhuai 古壞 and huguai 胡怪. The reading of 壞 [koi] is consistent with the fanqie 古 壞. The reading in the following characters can be explained by analogy: 果 (見 initial) > 踝 [koa], 塊傀 (溪 initial) > 槐 [koi], 貴 (見 initial) > 潰 [kuəi], 係 (見 initial) > 系 [kiəi], 溪 (溪 initial) > 蹊 [kiəi], 圭 (見 initial) > 畦 [kiu], 感 (見 initial) > 憾 [kam], 甘 (見 initial) > 酣 [kam], 押 (影 initial) > 狎 [ap], 甲 (見 initial) > 匣 [kap], 間 (見 initial) > 癇 [kan], 厷 (見 initial) > 紘 [koiŋ], 肱 (見 initial) > 宏 [koiŋ], 鬲 (見 initial) > 翮 [kiək] (Pak 1971: 44–46; Kōno 1979: 359–360). Ito (2007: 105) notes that the reading of 槲斛 [kok] and 莖 [kʌiŋ] may be a result of contamination. The reading of 槲斛 [kok] was probably contaminated from the initial of 角 [kak], and the reading of 莖 [kʌiŋ] may result from contamination of the initial of 經 (見 initial) [kiəŋ] or 輕 (溪 initial) [kiəŋ]. The yun 云 initial is generally represented as a zero-initial in EMSK. Historically speaking, the yun 云 initial originated from the xia 匣 initial. In Guangyun, the yun initial only appears before the third division and it is in a complementary distribution with xia. In other words, the yun initial is a palatalized xia in Qieyun. The representation in EMSK suggests that in the source language that SK borrowed from the yun initial had already occurred. The exceptional representations for the yun 云 initial are 彙 [hui], 鴞 [hio], 饁 [niəp], 燁 [niəp], 筠 [kiun] and 瑩 [hiəŋ]. The readings that result from analogy include 均 (見 initial) > 筠 [kiun], 熒 (匣 initial) > 瑩 [hiəŋ]. Pak (1971: 50) claims that 彙 [hui] is possibly from the analogy of 翬瀈 [hui]. This proposal, however, is subject to much debate, since 翬瀈 [hui] are not commonly used characters. It is contradictory to normal practice to derive the pronunciation of a high-frequency character from a low-frequency one. Lee (1997: 58) proposes that this reading may be a reflection of Old Chinese when the initial is [ɣ] or a reflection of the Old Mandarin form [huei]. Kōno (1979) notices that the reading of 彙 [hui] is [huei] in Modern Beijing dialect, and claims that the initial of 彙 [hui] was probably the same as the xia 匣 initial at the time of borrowing. Pak (1971: 50) proposes that the [h] in 鴞 ‍‍[hio‍] results from the analogy of the initial of 號 [ho]. The reading of 饁燁 [niəp] includes an extra [n] initial. The semantic notation before the 饁 [niəp] in HMJH is “[i pa tʌl],” which means “carry a meal to workers in field.” In YH the semantic notation before the 燁 [niəp] is “[pit nal],”

Initial system 79 which means “bright, flame.” A [n] initial was generated between the [l] coda in the preceding syllable and the following [i] due to the effect of assimilation (Nam 1973: 102–104; Lee 1997: 122).

3.5 Group duan 端 (alveolar stops, laterals and nasals) The duan 端 initial is mainly represented as [t]. The [th] reading in the following characters are believed to be a result of analogy: 㷟 (透 [tʰ] initial) > 鎚 [tʰoi], 𣨼 (透 [tʰ] initial) > 蔕 [tʰiəi], 超 (徹 [ʈh] initial) > 貂 [tʰio] (Pak 1971: 59). Pak (1971) states that 蔕 [tʰiəi] and 鎚 [tʰoi] result from the analogy of tou 透 initial characters 𣨼 [tʰiəi] and 㷟 [tʰoi], respectively. However, 𣨼 [tʰiəi] and 㷟 [tʰoi] are not high-frequency characters, so they are less likely to be the source of analogy. The reading of 妬 [tʰu] is probably contaminated from the initial of 拓 [thak] (Pak 1971: 59). The reading of 耽 [tʰam] may obtain the [tʰ] initial from the influence of deng 澄 initial character 沈 [tʰim] (Ito 2007: 64). The tou 透 initial is generally represented as [th] or, secondly, [t]. The [th] representation is probably a result of artificial effort to reflect the aspirated feature after the occurrence of aspirated [t] in Korean (Lee 1990: 90). The ding 定 initial is mainly represented as [t] or, secondly, [tʰ]. The reading of 恬 [niəm] may derived from the reading of the niang 娘 initial character 粘 [niəm] (Ito 2007: 76). In the MSK data we consulted, the character ting 挺 can only be found in YH, with a reading [niəŋ]. The [n] initial in 挺 [niəŋ] is probably a copier’s error. The ni 泥 initial is generally represented as [n] in EMSK. The reading of 拈 in MSK is lists as follows: HMJH [tsiəm], HD [tiəm], OP [niəm] (popular reading [tiəm]), SU, JU [niəm], YH [liəm]. The [tsiəm] in HMJH is probably derived from the analogy of the phonetic component 占 [tsiəm]. In the comparison in Table 3.5, the ni 泥 initial is represented as [l] in the following characters: 諾 [lak], 難 [lan], 赧 [lan], 寧 [liəŋ], 駑 [lo], 怒 [lo], all of which are from YH. The [l] representation of the ni 泥 initial can be explained by different transcription conventions. Native Korean phonology has a Dueum beopchik [Rule of initials], which prohibits the appearance of word-initial /l/ and regulates that a word-initial /l/ needs to change to [n] or should be dropped when it followed by /i/. The representation of the ni 泥 initial as [l] is related to the connection between [n] and [l] in the initial position. The lai 來 initial is generally represented as [l] in EMSK. The [n] representation of the lai 來 initial and the zero initial reading in 隷 [iəi] are a reflection of the reading after the application of Dueum beopchik [Rule of initials]. The reading of 巒 [man] is derived from the analogy of the ming 明 initial character 蠻 [man].

3.6 Group zhi 知 (retroflex stops and nasals) The zhi 知 initial is generally represented as [t] or [th] in EMSK. The [ts]/[tsh] representation of the zhi 知 initial can be attributed to the effect of palatalization. According to Eom (2003), the palatalization of the Middle Chinese zhi 知 initial

i

3

合

合 開

支 脂

iᴇ iᴇ w iᴇ i

w

ɣ

3 3B 3 3

開

支

ei

齊

4

開

祭

ɑi 帶 tʌi w ɑi jᴇi

1 1 3

開 合 開

泰

uoi tʌi 碓對 tʰoi 鎚堆

1

合

灰

jɤ jṷo̝ əi tʌi 戴 tai 歹

3 3 1

合 合 開

魚 虞 咍

tiəi 邸觝䏄帝堤羝 tʰiəi 蔕蝃嚏諦 tiə 低底詆柢抵

to 賭覩都 tu 蠹 tʰu 妬

uo

1

合

模

tʰa 𦖋

ɑ

w

1

合

戈

端 t ta 多

ɑ

開

歌

1

Table 3.5  The group duan 端 in EMSK 定 d ta 爹 tʰa 駄舵駝跎鼉

tʰiəi 涕體 tiəi 梯

tʰai 汰泰

tʰʌi 態胎 tʌi 貸 tʰoi 腿退 蓷煺

ti 地

tiəi 弟第悌娣鵜蹄啼稊綈荑 踢𧀰提踶 tʰiəi 髢棣締逮

tʰai 釱 tai 大

tʌi 袋黛代玳擡臺待岱tʰʌi怠苔 炲殆 tʰoi 頹㿉

tʰa 涶唾 tʰa 惰 墮 tʰai 妥 tʰo 土吐菟 to 肚鍍度屠徒途圖渡塗 兔 tu 杜

透 th tʰa 他拖

來 l la 羅蘿籮鑼

ni 泥

nai 奈柰

loi 癩瀨賴 loi 酹 liəi 例礪 liə 蠣厲勵 liəi 鱧醴禮棙儷荔犁盠鸝藜 liə 麗黎戾沴 iəi 隷i li 詈璃籬鱺離罹 li 羸 liu 累 li 履痢蜊梨 ni 利 liu 壘淚虆 niu 類

lo 櫓虜滷輅鷺路露艫蘆轤顱鱸纑鸕攎 爐𤬛 loi 賂 no艣 liə 侶膂旅濾鑢閭廬驢呂慮 lu 縷褸屢 nai 鼐乃耐 loi 賚 lʌi 麳來 nʌi 內 lʌi 耒 noi 餒 loi 儡擂雷罍磊

no 弩奴 lo 駑怒

na 懦糯捼 la 裸蓏瘰螺𥡜 lo 騾

泥 n na 挪

開

開

侯

尤 覃

帖

ep

em tiəm 店點

4

添

開

iᴇp

3

鹽

葉

ɑp

iᴇm

開

1

3 1

1

iᴇu eu tio 鳥釣雕錭彫弔凋 tʰio 貂 u tu 斗抖蚪𥆖陡 to 篼 tʰu 闘 ɨu əm tʰam 耽 tam 湛 əp tap 答 ɑm tam 膽擔

3 4

to 島擣祷刀舠倒到叨

ɨ ɑu

3 1

3

盍

開

開

開 開

宵 蕭

合 談

開 開

之 豪

tio 銚跳蜩鰷條調 tʰio 齠 tu 痘荳脰餖頭竇豆 tʰu 骰㢏䬦投

tʰiəm 舔添 tiəm 簟 㼭 tʰiəm 甜 忝 niəm 恬vi tʰiəp 帖貼 tiəp 蝶楪 tʰəp 牒 tʰiəp 疊

lu 耬瘻漏鏤簍螻髏摟蔞樓

li 鯉娌理里裏吏狸嫠 ni 李 lo 老䳓䇭澇撈𤛮勞 lio 潦轑醪 loi 牢 lio 燎療 lio 蓼鐐僚鷯聊遼了瞭寥 nio 料ii 陋

niəp 捻

niəm 鲇念 tsiəm 拈vii

liəp 鬛獵 niəp 躐v

(Continued )

liu 柳畱瘤溜餾琉旒榴鶹流留 nam 男南 lam 婪 楠 nam 嵐 nap 納衲 lam 攬爁覽纜濫籃襤 nam 藍iii lap 蝋 nap 臘鑞 liəm 臉殮𧸘廉匳簾帘鎌臁歛iv

niu 㝅

nio 尿

to 道稻盜桃陶萄淘濤逃導悼蹈 no 瑙腦 nio 臑

tham 探貪 tam 潭蟫壜藫覃 tam 黮 tap 錔踏 tam 毯菼 tam 噉淡談痰澹 tan 坍 thap 塔搭 榻塌

tʰu 偷透 tiu 鍮

tio 眺糶

tʰo 套討 thio 縧 to 饕韜

iᴇt en tiən 典巔癲顛

en et

ən uon ton 頓墩驐敦惇

uot tol 柮 in it

3 4

4

1 1

1 3 3

合 開

合 開

開 合

合 開

薛

先

屑

痕 魂

沒 真 質

w

w

ɣ

w

ɑt æn iᴇn w iᴇn iᴇt

末 刪 仙

2 3 3 3

ɑn tan 鍛碬端短斷

ɑt

w

tal 怛

ip ɑn

3 1

1

tan 担疽旦箪丹褝單

im

3

開 開 合 開

桓

合

開

緝 寒

曷

開

侵

Table 3.5 (Continued)

tʰʌn 吞 ton 褪暾

tʰiən 天 tʰiəl 餮鐵

tʰal 脫

tʰan 炭灘攤 擹嘆坦 tal 撻獺闥 韃 tan 湍猯

ton 沌囤豚魨遁 tun 臀飩屯鈍 tol 堗突

til 垤跌迭耋

tiən 殿澱電淀甸奠畋癜靛田塡 tin 殄

tʰal 奪

tan 段團椴

tʰan 袒彈誕憚 tan 壇檀 tal 薘達

nul 訥

non 嫩

niəl 捏 nal 苶

niən 撚 niən 年

lan 赧

nan 暖

lan 難

lin 悋𤗷麟鱗鄰吝 liul 栗篥

lon 論

liən 輦鰱連聯 liən 戀 liəl 蛚烈鴷冽 niəl 列ix裂 liəl 劣 liən 錬憐蓮鍊 niən 練x

lan 卵鸞亂 man 巒viii

lal 辣糲

lɨm 廩檁凜 lim 霖林臨 lip 粒笠立苙 lan 嬾瀾欄蘭懶爛

iaŋ

iak əŋ tɨŋ 鐙凳橙燈登等 ək tək 德 tɨk 得 ɨŋ ɨk ɣ æŋ iᴇŋ eŋ tiəŋ 頂鼎碇飣顁丁釘疔 訂 ek tiək 嫡 鏑靮蹢菂𢁕的滴

uŋ

juŋ uk

juk oŋ toŋ 冬

1

3

3 1 1 3 3 2 3 4

1

3 1

3 1

開

開

開 開 開 開 開 開 開 開

開

合

合

合

鐸

陽

藥 登 德 蒸 職 庚 清 青

錫

東

屋

冬

toŋ 蝀凍棟涷湩東

ɑk

1

開

4

in it ɑŋ taŋ 襠璫㼕黨讜當

w

w

3

合

諄 術 唐

tʰoŋ 統

tok 禿鵚

tʰiəŋ 廳聽 tiəŋ 汀鞓 tʰiək 趯剔 倜惕 tiək 逖 tʰoŋ 桶痛 通

tʰɨk 慝

nɨŋ 能

lak 諾

naŋ 曩囊

toŋ 疼彤

tok 牘犢髑 櫝讟瀆韣獨讀

toŋ 動峒衕 桐瞳童銅僮洞同

noŋ 膿農

noŋ 齈

tiəŋ 艇亭庭廷聤蜓定停 niəŋ 挺 niəŋ 佞 liəŋ 寧 tiək 荻敵笛狄糴覿迪 nik 溺 tʰiək 滌

tɨŋ 藤騰籐 tʰɨk 特

tʰaŋ 帑鐋湯 taŋ 塘糖堂棠螳膛螗煻唐 taŋ 儻 tʰaŋ 蕩 tʰak 托橐籜 tʰak 鐸度 飥𧘐拓 託 tʰik 驝xi

liuŋ 窿隆 lok 轆鹿麓漉磟簏𣿍 nok 祿 liuk 陸六戮

loŋ 籠聾櫳嚨䪊竉弄

liək 暦癧櫪靂䍽歷 lɨk 櫟礫

(Continued )

liaŋ 兩輛粮樑蜋梁涼粱 良亮量 liak 掠略 lɨŋ 稜 lɨk 肋勒鳓 lɨŋ 綾菱陵䔖凌 niək 力xiii lʌiŋ 冷 liəŋ 嶺領令 liəŋ 鈴伶零囹欞聆蛉鴒瓴翎靈

lak 駱落珞絡 nak 洛xii

liun 輪淪綸掄倫 liul 葎律 laŋ 浪狼廊螂稂朗

i ii iii iv v vi vii viii ix x xi xii xiii

ok tok 篤督 jṷo̝ŋ jṷo̝k [t]: 112 [th]:11 [th]:70 [t]:25

[t]: 161 [th]:36 [n]:2

tok 纛碡毒

[n]:42 [l]:6 [ts]:1

loŋ 壟龍 lok 綠錄 [l]: 291 [n]:16 [ø]:1 [m]:1

HMJH, CJM [iəi], HD, SU, JU [liəi] YH [nio] YH, CJM [nam], SG, HD, SU, JU [lam] 歛 [liəm] is a xiao 曉 initial tan 談 rhyme character. In HMJH, 歛 [liəm] is used as a variant of 斂 [liəm], which means “to collect.” YH [niəp] YH, SE [niəm], CJM [liəm], HD, OP [thiəm] (popular reading [niəm]) HMJH [tsiəm], HD[tiəm], OP[niəm] (popular reading [tiəm]), SU, JU [niəm], YH [liəm] YH, YJ [man] YH, SE [niəl], IG, SH, CJM [liəl] YH [niən] HMJH (YS edition) [tʰik], HMJH (DD edition) [tʰak] CJM, SE [nak] YH, SE, CJM (SB edition) [niək], CJM (GJ edition) [nək], NE, SG, JY, SH [liək]

沃 合 1 鍾 合 3 燭 合 3 Summary

Table 3.5 (Continued)

æ jɤ

jṷo̝

ɛ æi jᴇi wi ᴇi iᴇ

iᴇ

2 3

3

2 2 3 3 3

3B 3

3 3

合 合

合

開 開 開 合 開

開

合 開

開

開

魚

虞

佳 夬 祭

脂

之

肴

宵

支

æ

2

開

麻

iᴇu

3

tio ‍朝‍

tio ‍嘲‍‍

æu

ɣ

tʰiu 追 tʰi 置

tʰi 致

tʰiu 綴 ti 智蜘知

tsoa 檛 tiə 瀦貯豬 tiəi 豬 tiu 拄株蛛誅 tsu 註

i

2

ɨ

w

i

ɣ

ɣ

ɣ

wɣ

ɣ

知 ʈ tʰa 打

Table 3.6  The group zhi 知 in EMSK

tʰio 超

tʰi 癡恥 tsi 祉 tʰʌi 笞

tʰi 絺

tsʰi 黐

tʰai 蠆

tiə 楮樗摴 tʰə 攄

徹 ʈh tʰa 詫

tio 朝‍潮肇趙兆 sio召‍‍ii

to ‍棹‍‍

tʰiəi 馳 ti 池篪 tʰi豸 tʰiu 錘 tʰi 雉稚緻 ti 墀遲 tʰoi 椎 tʰiu 墜 tʰi 痔峙治值 ti 持

tʰiəi 彘滯

tiə 苧杼筯儲 tiəi 除 tiu 柱幮廚住

澄 ɖ ta 茶 tsʰa 搽

nio ‍鬧‍‍ io‍ 鐃‍‍i

ni 你

ni 尼

nai 嬭

niə 女

娘 ɳ

(Continued )

ɛm ɛp iᴇm

iᴇp

im

ip ɣ ɛn

iᴇn w iᴇn iᴇt w iᴇt in it w in w it iaŋ

iak w iak

3

3

3

3 2

3 3 3 3 3 3 3

3

3

開

開

開

開 合 開 合 開

開

葉

侵

緝 山

藥

真 質 諄 術 陽

薛

仙

開 合

合

ɣ

ɣ

2

開

咸 洽 鹽

ɨu

開

尤

3

Table 3.6 (Continued)

tiən 𩥇展 tsiən 鳣 tiən 轉 tʰiəl 哲 tʰiəl 啜綴掇輟 tin 珍鎮 til 銍窒 tun 窀迍 tsul 茁 tiaŋ 帳張 tʰiaŋ 脹漲 tʰiak 着

tʰim 砧 sim 椹

tsʰam 站 tsap 劄 tiəm 沾

tiu 肘晝輈

tsʰiun 椿 tʰiul 黜 tʰiaŋ 韔悵暢

tin 趁

tʰiəl 徹

tʰiəm 諂 tiəm 覘

tʰiu 抽

tʰiak 著

tʰiul 𦬸朮 tiaŋ 丈杖腸場長

tin 塵陳 til 姪帙秩

tiən 廛纏 iən 椽 tiən 傳 tʰiəl 轍澈

tʰip 蟄 tʰan 綻

tʰim 沈

tam 賺

tiu 宙胄紬籌疇儔稠

niaŋ 娘孃 iaŋ 釀

nil 昵

niən 碾

nim 賃

niəp 鑷

niu 紐鈕扭糅 liu 杻狃

juk

jṷo̝ŋ jṷo̝k

3 3

3

3 3

開 合

合

昔 東

屋

鍾 合 燭 合 Summary

[t]: 35 [th]:25 [ts]:6 [tsh]:1 [s]:1

tiuŋ 中 tʰiuŋ 忠衷 tiuk 竹 tʰiuk 築 tʰioŋ 塚

tiək 謫摘 tiəŋ 貞

tsoaŋ 樁 taŋ 戇 tʰak 卓 啄椓瘃琢斵 tiŋ 徵 tʰiək 陟

[th]:24 [t]:9 [ts]:1 [tsh]:2 [l]:1 [p]:1

tʰioŋ 寵

tʰiuk 蓄畜

tiəŋ 蟶檉赬偵 liəŋ 逞iii piŋ 騁iv

tʰik 鶒敕飭 tʰʌiŋ 撐 tʰʌik 坼

tʰak 踔

HMJH [io], HD, SU, JU [nio] YH, SE, SG [sio], HD, OP [tsio] (popular reading [sio]) YH [liəŋ], HD [tsiəŋ] (popular reading [thiəŋ]), OP [tsiəŋ] (popular reading [liəŋ]) YH [piŋ], HD, OP [tshiŋ] (popular reading [piŋ])

iᴇk juŋ

2 2 2 3

開 開 開 開

i ii iii iv

ɨŋ ɨk ɣ æŋ ɣ æk ɣ ɛk iᴇŋ

ɣ

ɔk

ɔŋ

ɣ

3

2

開

開

蒸 職 庚 陌 麥 清

覺

江

tiuŋ 重 tʰiok 䕽躅 [t]: 55 [th]: 40 [tsʰ]:1 [ø]:1 [s]:1

tʰiuk 軸逐柚舳妯

tʰiək 擲𧃐 tiək 躑 tʰiuŋ 沖蟲翀 tiuŋ 蚛仲

tiəŋ 呈裎程

tɨŋ 澄 tiŋ 懲 tik 直 tiəŋ 鋥 tʰʌik 宅澤擇

tʰak 濁濯擢

taŋ 幢撞

[n]:16 [l]:4 [ø]:2

noŋ 醲 loŋ 濃

lik 匿

88  Initial system was noticeable in SK since 1448 and seems to have been completed by 1751 at the latest. The [ts]/[tsh] representation shows that this palatalization has occurred in some characters at that time. The reading of 椹 [sim] is derived from the analogy of its phonetic component 甚 [sim] (Pak 1971: 59). The che 徹 initial is generally represented as [th] or [t] in EMSK. The [tsh]/[ts] representation of the che 徹 initial arises from palatalization of the initial. The reading of 椿 [tsʰiun] can also be explained by analogy: 春 (昌 [tɕʰ] initial) > 椿 [tsʰiun]. The reading of 逞 [liəŋ] was probably [tiəŋ] in the beginning, and became [liəŋ] as the weakness of the initial from [t] to [l] (Kōno 1979: 375). The reading of 騁 [piŋ] is derived from analogy of pang 滂 initial character 聘 [piŋ]. The deng 澄 initial is represented as [t] or [th] in EMSK. 搽 [tsʰa] is a reflection of the Old Mandarin reading [tʂʰa] as shown in Zhongyuan yinyun 中原音 韻 (1324) (hereafter ZYYY). There are two fanqie spellings for zhao/shao 召 in Guangyun, which are zhizhao 直照 and shizhao 寔照. The reading of 召 [sio] is consistent with the fanqie shizhao 寔照. The reading of 椽 [iən] is probably derived from the analogy of the high-frequency character 緣 [iən]. The niang 娘 initial is generally represented as [n] in EMSK. The representation of niang 娘 initial with [l] in some characters is related to the connection between [n] and [l] in the initial position due to Dueum beopchik [Rule of initials]. The reading of 鐃 [io] results from analogy of its phonetic component 堯 [io]. The reading of 釀 [iaŋ] is the same as ri 日 initial characters 禳蘘 [iaŋ] and the xin 心 initial character 驤 [iaŋ], showing the effect of analogy based on phonetic component.

3.7 Group jing 精 (alveolar fricatives and affricates) The jing 精 initial is represented as [ts] or [tsʰ] in EMSK. The readings that derived from analogy include 纖 (心 initial) > 殲 [siəm], 俠狹挾 (匣 initial) > 浹 [hiəp]. The mainstream representation of the qing 清 initial is [tsh], although [ts] reading can be found in some characters. The reading of 焠 [siu] derived from analogy: 粹 (心 initial) > 焠 [siu]. The reading of 猜 [sɨi] cannot be explained by analogy. Ito (2007: 92) suggests that the initial of 猜 in the source input could be [s] given that the reading of 猜 is [sa] in Sino-Vietnamese. The cong 從 initial is represented as [ts] or [tsʰ] in EMSK. The reading of 蕈 [sim] is derived from the analogy of shu 書 initial characters 㜤瞫 [sim] (Pak 1971: 80). The reading of 吮 [iən] results from analogy of the yi 以 initial character yun 允 [iən]. The xin 心 initial is generally represented as [s]. The readings that derived from analogy are listed as follows: 緌 (日 [ȵ] initial) > 綏 [iu], 澡繰璪 (精 [ʦ] initial) > 噪臊燥 [tso], 準 (章 [tɕ] initial) > 隼 [tsiun], 悄 (清 [ʦʰ] initial) > 綃 [tsʰio], 參 (清 [ʦʰ] initial) > 糝 [tsham], 俊 (精 [ts] initial) > 峻浚 [tsiun], 怵䟣 (徹 [ʈʰ] initial) > 訹 [tʰiul], 壤攘穰 (日 [ȵ] initial) > 驤 [iaŋ], 剔惕逷 (透 [tʰ] initial) > 裼 [tʰiək], 茸 (日 [ȵ] initial) > 聳 [ioŋ] (Pak 1971: 82–83; Kōno 1979: 391–392). The reading of 恤 [hiul] obtains the initial from its phonetic component 血 [hiəl] (Lee 1990: 84; Kōno 1979: 391–392). The xie 邪 initial is represented as [s] in EMSK.

開

合

合

合

開

合

開 合 開 合 開

麻

模

魚

虞

咍

灰

泰

齊

祭

開 合

歌 戈

ɑ w ɑ

jæ

uo

jɤ

jṷo̝

əi

uoi

ɑi w ɑi jᴇi w jᴇi ei

1 1

3

1

3

3

1

1

1 1 3 3 4

tsiəi 霽濟擠韲虀齎 tsʌ 泲

tsʰoi 最 tsiəi 祭穄際

tsʌi 宰𣪮載栽再災哉

tsʰiui 娵

tsiə 蛆苴

tsiə 姐 tsiə 罝 tsʰia 借 tsʰa 嗟 tso 祖租組

精 ts tsoa 左佐

Table 3.7  The group jing 精 in EMSK

tsʰiəi 砌 tsʰiə 妻凄

siu 焠 tsʰoi 催淬 tsʰoai 蔡

tsʰo 醋 tso 措 tsʰu 麤 tsiə 疽沮鴡 tsʰiə 覷 tsʰu 趨 tsʰiui 取娶 tsʰʌi 綵菜彩採sɨi猜iv

tsʰia 且

清 tsh tsʰa 搓磋

寫iii瀉卸

siu 鬚須

siə 絮胥

so 訴嗉素塑蘇酥櫯遡泝

sia

心 s sa 挱 soa 鎖 sa 梭蓑 soai 瑣

tsiəi 薺臍蠐齊

siəi 歲 siəi 洗細 siə 壻西棲犀 sɨi 嘶

tsʌi 裁財才材纔在 sʌi 塞賽 sɨi 顋 tsoi 罪 tsʰoi 摧 soai 碎

tsiu 𡲰 tsiə 咀 tsʰiui 聚

tso 祚粗

tsia 褯ii藉

從 dz tsa 鹺咱 tsoa 坐矬i tsʰa 銼

(Continued )

siə 嶼芋序漵叙 緖徐

sia 斜邪謝

邪 z

開

開

開

開 開 開

脂

之

豪

宵

蕭 侯 尤

覃 合 談

開

iᴇp em

1

開

3 4

əm əp ɑm

1

開

葉 添

eu u ɨu

4 1 3

ɑp iᴇm

iᴇu

3

3

ɑu

1

盍 鹽

ɨ

3

w

i

i

iᴇ

3 3 3

合 開 合

w

iᴇ

3

開

支

Table 3.7 (Continued)

tsʰiəm 尖 siəm 殲xi tsiəp 椄接 tsʰiəp 睫 tsʰʌm 僭

tsam 揝 tsap 咂匝

tsu 奏走 tsiu 酒 tsʰiui 僦

tsʌ 子滋孳耔镃鼒兹 tsʌi 梓 tso 藻‍早 蚤澡棗 𥨫糟躁遭 tsʰio‍‍ 椒‍焦‍礁‍蕉鷦醮

tsʰiuiəi 嘴 tsʌ 姊資姿恣諮 tsʌi 齍 tsʰiui 醉

精 ts tsʌ 紫髭

tsʰiəp 妾

tsʰiəm 籤憸

tsu 腠湊 tsʰiu 秋 鰍楸湫 鞦鶖䐐鞧 tsʰam 參慘

tsʰio‍‍ 锹‍

tsʰo‍ 草‍騲 ‍tso 糙操

tsʰʌ 次 tsʰui 翠

清 tsh tsʌ 莿雌刺 tsʰʌ 此

tsʌm 潛 tsiəm 漸 tsʰiəp 捷

tsʌm 蠶 tsap 雜 tsʰam 慙 tsam 暫

tsʰiu 鷲 tsʰiui 就

tsʌ 字牸慈 鶿 tso 皂‍造 ‍槽螬 曹‍艚 tsʰio‍‍ 誚‍樵憔

tsʌ 自瓷餈 tsʰiuiəi 悴萃

從 dz tsʌ 骴疵 tsʰi 漬

sap 𢻨 siəm 纖

tsham 糝x sap 靸 sam 三

心 s sʌ 璽賜撕𪆁斯徙 sɨi 廝 siu 髓 sʌ 死四私肆 siu 荽邃粹 iu 綏v si 葈 sʌ 笥司鸶蕬絲思伺 su‍ 嫂艘 ‍so‍ 搔繅騷掃 tso‍ 噪vi臊vii燥viii sio‍ 笑宵‍霄焇鞘小消銷逍 tsʰio‍ 綃‍ix sio 嘯簫蟏蕭 su 擞瞍叟薮嗽漱 siu 繡銹饈脩秀羞

siu 袖岫泅囚

siu 繸燧篲穗 遂 sʌ 姒耜祀飼寺 祠詞辭嗣似

siu 隨

邪 z

ep im ip

ɑn ɑt w ɑn

ɑt iᴇn

iᴇn

w

w

iᴇt w iᴇt en

et uon

uot in it

4 3 3

1

1

1 3

3

3 3 4

4 1

1 3 3

開

合

合 開

合

寒 曷 桓

末 仙

開 合 開

開 合

合 開

先

屑 魂

沒 真 質

薛

開 開 開

帖 侵 緝

tsiən 牮箋 tsʰiən 薦 tsiəl 癤節 tson 尊 tsun 樽 tsol 卒𣨛 tsin 進津晉

tsiən 鐫

tsiən 剪箭煎湔翦

tsʰan 讚贊 tsʰal 拶 tsʰan 鑽 tsan 纂

精 ts hiəp 浹xii tsʰim 浸

tsʰin 親 tsʰil 七漆㯃

tsiəl 竊切 tsʰon 寸村

tsʰiən 蒨千𢶊

tsiən 痊悛

tsʰal 撮 tsʰiən 韆遷淺

tsʰan 爨鑹

tsʰan 粲飡

tsʰim 寢䈜侵 tsɨp 緝

清 tsh

tsin 螓秦盡 tsil 疾嫉

tsiəl 截 tson 鐏存 tsun 蹲

tsiəl 絶 tsiən 前

tsʰiən 泉 iən 吮xiv tsiən 全

tsiən 餞錢 tsʰiən 踐賤

sim 蕈 tsip 集 tsɨp 楫輯 tsan 殘

從 dz

sin 信訊顖辛薪新迅 sil 蟋悉 sɨl 膝

siəl 糏楔屑 son 孫猻蓀飧損巽遜

siəl 疶泄褻 siəl 雪 siən 跣霰先

siən 宣選

siən 癣燹藓獮線仙秈

san 算酸狻餕 suən 蒜

san 傘糤珊散

心 s siəp 爕 sim 心

(Continued )

sin 燼

siən 旋璇

iən 涎xiii siən 羨

sim 燖鬵尋 sɨp 習襲

邪 z

開 開

開 開

開

開

開 開

藥 登

德 蒸 職

清

昔

青 錫

合

開

陽

東

開 開

合

唐 鐸

術

諄

iᴇŋ

iᴇk

eŋ ek

uŋ

juŋ

3

3

3

4 4

1

3

iak əŋ

3 1

ək ɨŋ ɨk

iaŋ

3

it

w

ɑŋ ɑk

in

w

1 1

3

Table 3.7 (Continued)

tsoŋ 鬉鯼 tsʰoŋ 緫

tsiək 勣績

tsʰɨk 則xviii tsɨŋ 甑 tsik 稷 tsɨk 鯽即 tsiəŋ 井精旌 tsʰiəŋ 菁睛 tsiək 襀跡積 tsʰiək 脊鶺蹐

tsiaŋ 醬將漿槳獎 tsiak 雀爵 tsɨŋ 罾矰增憎

tsaŋ 葬贓臧 tsak 作

精 ts tsiun 俊儁 tsun 遵

tsʰoŋ 葱囪聰怱

tsʰiək 刺 tsiək 磧 tsʰiəŋ 青鯖蜻 tsʰiək 戚鏚慼

tsʰiəŋ 清圊凊

tsʰaŋ 槍搶 tsiaŋ 鏘 tsiak 鵲

tsʰaŋ 鶬倉蒼 tsʰak 錯

清 tsh tsiun 皴逡竣

tsʰoŋ 叢

tsiək 寂

tsiəŋ 穽情靜淨靖 tsʰiəŋ 晴請 tsiək 籍 tsʰiək 瘠塉

tsaŋ 欌臟藏 tsak 柞昨 tsʰak 鑿 tsiaŋ 匠樯薔 嫱牆 tsiak 嚼 tsɨŋ 贈 tsʰɨŋ 層 tsək 賊 tsɨŋ 繒

tsol 崒

從 dz

sioŋ 菘𪀚

siəŋ 星鯹腥醒惺 siək 錫淅蜥析 tʰiək 裼xix soŋ 送

siək 昔澙磶腊舃惜

siəŋ 姓性省

sik 息熄

sʌik 塞

siaŋ 鮝相箱廂緗想 iaŋ 驤xvii siak 削 sɨŋ 僧

心 s siun 筍詢 tsiun 隼峻浚 siul 銊 hiul 恤xv tʰiul 訹xvi saŋ 顙嗓磉𤸯桑喪 sak 索

siək 夕席 汐穸

siəŋ 餳

siaŋ 象橡 翔庠詳祥

邪 z siun 旬巡循徇馴

jṷo̝k tsiok 足 [ts]: 115 [tsh]:29 [s]:1 [h]:1

3

合 合

冬 鍾

燭 合 Summary

tsʰok 促 [tsʰ]: 81 [ts]: 23 [s]:2

清 tsh tsok 簇

[ts]: 84 [tsh]: 23 [ø]:1 [s]:1

tsioŋ 從

從 dz tso k族

soŋ 駷 sioŋ 竦悚 ioŋ 聳xx sok 粟 [s]:190 [ts]:6 [tsh]:2 [th]:2 [h]:1 [ø]:3

心 s sok 速 siuk 夙蓿肅宿

siok 俗續 [s]: 62 [ø]:1

sioŋ 頌誦訟松

邪 z

i HMJH (YS edition) [tsoa], HMJH (DD edition) [tsʰa] ii HMJH (YS edition) [tsia], HMJH (DD edition) [tsʰia] iii The Sino-Korean reading of xie 寫 is [s*ia] in HMJH (YS and DD edition) and [sia] in CJM, YH and SG. Synn (2006: 65) notes that “in HMJH, the native Korean reading under the entry xie 寫 is [s*ɨl]. The [s*] initial in [s*ia] is probably derived from the [s*] in the native Korean reading [s*ɨl].” iv YH [sɨi], HD [sai] (popular reading [sɨi]), OP [tshʌi] (popular reading [sɨi]) v YH, CJM, SE, SG [iu], HD [ziui] (popular reading [ziu]), SU [siu] vi HMJH [tso], HD, OP [so] (popular reading [tso]), SU, JU [so] vii HMJH [tso], HD, OP [so] (popular reading [tso]), SU, JU [so] viii YH [tso], IG [tso], OP [so] (popular reading [tso]) HMJH, YH [tshio], HD, OP [sio] (popular reading [tshio]), SU, JU [sio] ix x HMJH (YS edition) [tsham], HMJH (DD edition) [sʌm] xi YH, SE, SG, HD [siəm], OP [tsiəm] (standard reading [siəm]) xii YH, HD [hiəp], OP [tsiəp] (standard reading [hiəp]) xiii HMJH, YH, HD [iən], OP [siən] (standard reading [iən]), SU, JU [siən] xiv HMJH [iən], HD, SU, JU [tsiən] xv HMJH [hiul], HD, OP [siul] (popular reading [hiul]) xvi HMJH [tʰiul], HD, SU, JU [siul] xvii HMJH, CJM [iaŋ], HD, OP [siaŋ] (popular reading [iaŋ]), SU, JU [siaŋ] xviii In addition to [tsʰɨk], YH also lists the reading [tsɨk]. SH [tsʰɨk], CJM, NE, SE, IG, SG, JY, OP [tsɨk], HG [tsʰɨk]/[tsɨk] xix HMJH [tʰiək], SG, HD, SU, JU [siək] xx YH [ioŋ], HD, OP [sioŋ] (popular reading [ioŋ])

uk juk oŋ tsoŋ 綜宗 jṷo̝ŋ tsioŋ 蹤縱

1 3 1 3

合

屋

精 ts tsok 鏃

94  Initial system The reading of 涎 [iən] is derived from the analogy of its phonetic component 延 [iən].

3.8 Group zhuang 莊 (retroflex fricatives and affricates) The zhuang 莊 initial is generally represented as [ts] or [tsʰ] in EMSK. In the following three characters 詐 [sa], 渣 [sa] and 𨧫 [sʌ], the zhuang initial is represented as [s]. The reading of 𨧫 [sʌ] is derived from the analogy of its phonetic component 事 [sʌ]. In the YS edition of HMJH, the zhuang initial is represented as [k] in 抓 [koa], a reading possibly derived from the analogy of the character 瓜 [koa]. The notation for 抓 [koa] was changed to [tsoa] in the DD edition of HMJH. The chu 初 initial is generally represented as [tsʰ]. In the following three characters, 鍤 [sap], 插 [sap] and 鏟 [san], the chu initial is represented as [s]. The reading of 鍤插 [sap] may result from the analogy of 歃喢 [sap] (Pak 1971: 82). The reading of 鏟 [san] is probably derived from the analogy of the shan 山 initial character 產 [san]. The cong 崇 initial is represented as [s], [ts] or [tsʰ]. It is not easy to conclude which reading is the representative form. Kōno (1979: 402–403) claims that the [s] reading is mainly seen in the vernacular reading, and thus should be the older layer, while the [ts] and [tsh] are a new layer influenced by Chinese at a later time. The sheng 生 initial is generally represented as [s]. The readings of the following characters result from analogy: 债 (莊 [tʂ] initial) > [tsʰai], 衰 (初 [tʂʰ] initial) > 榱 [tsʰoi], 峭哨陗 (清 [ʦʰ] initial) > 梢艄稍 [tsʰio] (Pak 1971: 82). The reading of 縮 [tsʰiuk] may derived from the analogy of 蹙 [tsʰiuk], which carries a similar meaning with 縮 (Ito 2007: 94).

3.9 Group zhang 章 (palatal fricatives and affricates) The zhang 章 initial is mainly represented as [ts] or [tsh] in EMSK. The exceptional representations that result from analogy include 召 (禪 [dz] initial) > 昭沼 [sio], 超 (徹 initial) > 招 [tʰio], 窒銍 (知 initial) > 蛭桎 [til], 屯鈍 (定 [d] initial) > 肫 [tun] (Pak 1971: 78–79; Lee 1990: 150). There are two fanqie spellings for 診 [tin] in Guangyun, which are zhangren 章忍 and zhiren 直刃. The reading of 診 [tin] is consistent with the second fanqie spelling. There are three fanqie spellings for 缜 in Guangyun, which are chouren 丑人, changzhen 昌眞 and zhangren 章忍. The reading of 缜 [tin] is consistent with the first fanqie spelling. Li (1985: 38) proposes that the reading of 疹畛 [tin] is a reflection of Old Chinese, in which the zhang 章 initial was dental. Lee’s proposal is reasonable, but it is necessary to point out that the reading of 疹畛 [tin] could be a result of analogy: 珍 (知 initial) > 疹畛 [tin]. The chang 昌 initial is represented as [tsʰ] or [ts] in EMSK. The following representations are derived from analogy: 迢 (定 initial) > 弨 [‍tʰio]‍‍‍, 審 (書 initial) > 瀋 [sim], 叔 (書 initial) > 俶[siuk]. The reading of 踳 is [siun] in YH, and [tshiun] in HD. The reading of 踳 [siun] in YH may derived from analogy, though the source of analogy is not clear.

ɛ

æi iᴇ w iᴇ i w i

ɨ

ɣ

ɨu

ɛm ɛp

ɣ

2

2 3 3 3 3

3

2

3

2

開

開 開 合 開 合

開

開

開

開

佳

夬 支

之

肴

尤

咸 洽

脂

æ jɤ jṷo̝ ɣ ɛi

魚 虞 皆

ɣ

ɣ

ɣ

æu

wɣ

2 3 3 2

合 合 合 開

æ

ɣ

2

開

麻

tsʰiu 甃緅 tsʰu 皺縐 tsam 蘸 tsʰam 斬

tsʌi 滓 tsʰɨi 輜鲻 sʌ 𨧫 tso 爪‍笊 koa‍ 抓‍‍i‍

tsʰai 債跐

tsʌi 𢋿齋

tso 葅阻 tsiə 詛

莊 tʂ tsa 鮓蚱炸榨皶 tsia 樝 tsʰa 齇𤹡 sa 詐渣

Table 3.8  The group zhuang 莊 in EMSK

sap 鍤插

tsʰu 篘

tsʰio 炒‍鈔‍

tsʰi 廁

tsʰiuiəi 揣

tsʰa 釵

tsʰo 礎初楚 tsʰu 芻

初 tʂʰ tsʰa 岔汊靫杈叉差

su 愁 tsʰiui 驟 tsʰam 讒饞 tsap 牐煠

so‍ 巣

si 柿 sʌ 士事俟仕

tsʰai 寨

sə 鋤 tso 助 tsʰu 雛 tsiəi 儕 sɨi 豺 sɨi 柴

崇 dʐ sa 槎乍查

(Continued )

sam 釤杉 sap 歃箑

‍so 鞘 sio 弰‍蛸‍筲 ‍ʰio 梢‍艄稍 su 瘦蒐溲螋搜

sʌ 師獅螄 siu 帥 tsʰoi 榱 soi 衰 sʌ 史使

sɨi 簁釃

tsʰai 𨢦 soai 灑

soa 耍 so 所䟽蔬梳踈 su 數

生 ʂ sa 沙裟紗鯊

ip

ɣ

ɛn

ɛt æn wɣ æn ɣ æt w ɣ æt w iᴇn in ɪn ɪt w it iaŋ

ɔŋ ɔk ɨk

æŋ

3

2

2 2

2 2 3 3 3 3 3 3

2 2 3

2

開

開 開 合 開 合 合 開 開

緝

山

黠 刪

合 開

開 開 開

開

仙 真 臻 櫛 術 陽

江 覺 職

庚

鎋

ɣ

ɣ

ɣ

ɣ

ɣ

æm im

開 開

ɣ

銜 侵

2 3

Table 3.8 (Continued)

tsʰak 捉 tsʰɨk 仄側 tsʰʌik 昃 tsʌiŋ 爭

tsaŋ 粧裝莊壯

tsiən 跧 tsin 臻 tsin 榛 tsɨl 櫛

tsan 盞剗

tsʌm 簪 tsʰʌm 譖 tsɨp 蕺戢

莊 tʂ

tʰʌiŋ 鐺

tsʰɨk 廁测惻

tsʰaŋ 窗

tsʰaŋ 瘡創

tsʰin 齔櫬

tsʰal 剎

tsʰal 察 san 鏟

初 tʂʰ

tsʌiŋ 崢

tsʰak 鋜

tsaŋ 狀 saŋ 牀床

tsʰan 饌撰

tsʰal

tsan 棧孱

tsʌm 涔梣

崇 dʐ

sʌiŋ 生甥笙鉎鼪 省眚 siəŋ 猩

saŋ 瀧 soaŋ 雙 sak 朔槊嗍 sʌik 穑嗇色

sɨl 瑟蝨 sol 蟀率 saŋ 霜孀爽

soa 刷

sal 殺 san 訕刪 soan 孿𣟴

san 產𩥮山疝

sɨp 澁

生 ʂ sam 衫 sʌm 蔘森滲

合 合

東 屋

ɣ

æk ɣ ɛŋ ɣ ɛk

juŋ juk

2 2 2

3 3 [ts]:32 [tsh]:20 [s]:3 [k]:1

莊 tʂ tsʰʌik 笮窄 tsʌiŋ 箏諍 tsiək 幘 tsʰʌik 謮責

i HMJH (YS edition) [koa], HMJH (DD edition) [tsoa], HD, SU, JU [tso]

Summary

開 開 開

陌 耕 麥

[tsʰ]:30 [s]:3 [tʰ]:1

tsʰʌik 策冊

初 tʂʰ tsʰʌik 柵

[s]:16 [ts]:10 [tsʰ]:9

siuŋ 崇

崇 dʐ

[s]:72 [tsʰ]:6

tsʰiuk 縮

生 ʂ

iᴇ

w

i

w

ɨ

iᴇu

3

3

3

3

3

合

開

合

開

開

之

宵

脂

i

iᴇ

3

開

支

jᴇi

w

3

合

祭

jᴇi

3

開

祭

jṷo̝

3

合

虞

jɤ

3

合

æ jæ

魚

ɣ

開 開

麻

2 3

‍‍tsio 照詔𨹸 sio ‍沼vii昭viii tʰio 招ix

tsʰiu 萑 tsʰio 隹錐 tsi 址沚趾痣志芝 止之

tsi 旨指砥脂𧹛至祗

ts i紙咫肢枝搘梔支 tsʰi 巵 tsʰiuiəi 箠捶

tsʰiuiəi 贅

tsu 炷注鑄 tsiu 主蛀朱珠 tsiəi 制製

tsia 赭柘炙者 tsʰia 遮 tsiə 渚 tsia 煮翥 tsiəi 諸

章 tɕ

Table 3.9  The group zhang 章 in EMSK

tsʰi 齒蚩 幟熾 嗤 tsʰio ‍麨 ‍tʰio ‍弨‍‍x

tsʰiu 推

tsʰiui 炊 tsʰiu 吹 tsʰi 痓鴟

tsʰi 眵侈

tsʰiui 樞

tsʰiə 處 tsiə 杵

tsʰia 車

昌 tɕʰ

sio 紹邵劭

si 市䦙時蒔 塒侍恃

si 鍉匙氏 是豉 siu 睡垂 siə 瑞 si 視 ki 嗜vi siu 誰脽

siu 樹銖殊 豎 siəi 誓 siə 噬筮逝

siə 墅署曙 薯

sia 社

禪 dʑ

sio 燒少‍

si 詩始試

siu 水

si 矢屎屍鳲蓍

siəi 世貰 siə 勢 siəi 稅帨 iəi 蛻ii si 豕翅絁施 i弛iii

siu 輸

書 ɕ sia 舍賒 sia 奢捨 sa 赦 siə 鼠暑黍𧑓恕書 鵨庶舒

si 示

tiəi 䑛iv

sia 射i麝 蛇 siə 抒紓

船 ʑ

zi 耳栮餌 珥㮌輀 i而 io 蕘xi饒擾 繞遶 zio 橈

ziu 緌

zi 樲二貳

zi 爾邇 zʌ 兒 iəi 蘂v

ziu 乳孺儒 iu 襦濡

ziə 汝洳 iə 茹如

日 ȵ

io‍ 曜‍鷂‍窰‍瑤‍軺‍謠‍舀 遙搖耀

i 苡頤飴貽苢以 異怡

iu 壝惟遺唯維

i 夷姨

i 匜椸扅移迤

iu 諭諛榆逾庾愈 愉踰裕 iəi 勩裔 thiəi 靾 iəi 銳睿

iə 蕷予蜍鸒輿餘余 與藇 iəi 譽預豫

ia 夜㖡野冶爺也

以 j

iᴇm

iᴇp

im

ip

iᴇn

iᴇn

w

iᴇt w iᴇt

in

it

3

3

3

3

3

3

3 3

3

3

開

合

葉

侵

緝

仙

開

真

質

開 合

薛

開

開

鹽

ɨu

3

開

尤

tsil 礩櫍質 til 蛭xx桎xxi

tsiəl 靼折 tsiəl 棁 tsol 拙 tsin 賑真振震 tin 疹畛診缜

tsiən 甎專剸

tsiən 戰氈饘鹯

tsip 執

tsim 斟 tsʰim 枕鍼 tsʌm 箴

tsiu 洲州舟周賙 tsʰiu 箒 tsʰiuk 呪xii tsiəm 占 tsʰiəm 瞻

tsɨl 叱

tsʰiən 喘 釧穿 川舛

tsʰiən 闡

sim 瀋xvi

tsʰiəm 韂

tsʰiui 醜 臭

sin 腎臣晨 宸辰鷐慎 siun 蜃

sip 十 sɨp 拾 siən 善鱔蟮 蟬饍禪 tsʰiən 擅 tsʰiən 腨篅

tʰim 忱xvii sim 甚

siəp 涉

siəm 蟾贍

siu 受壽售 酬讎綬

sil 室實失

sin 哂身紳娠伸申

siəl 設 siəl 說

siən 騸扇 tsiən 羶

sɨp 濕

sim 嬸深審

siəp 韘攝

siəm 苫閃 tsiəm 痁xiii

siu 手首狩獸收守

sin神

siəl舌

siən 船

sim 葚

zin 訒刃仁 人 in 忍仞認 zil 日

ziən 蠕 㮕撋 iən 軟 iəl 熱 siəl 𤑔xix

iən 然

zim 噛荏餁 絍任妊 im 紝 nim 恁稔xviii ip 入

ziəm 染髯 枏

ziu 葇 iu 柔揉

(Continued )

il 溢逸泆鎰

in 蚓引寅 iun 胤

iəl 悅閱

iən 鳶鉛沿緣捐

iən 蜒筵演延衍

ɨm 淫

iu 牖莠誘柚游蚰 油蝣斿猶由猷 輶攸 iəm 燄扊閻鹽艷 焰豔 ziəm 𢸴xiv tsʰiəm 簷xv iəp 葉

開

蒸

東

昔

清

iᴇk

iᴇk

w

juŋ

3

3

3

開

合

合

iᴇŋ

w

3

ɨk

tsioŋ 終螽 tsiuŋ 眾

tsiək 炙 tsʰiək隻

tsiəŋ 政 鉦𢁿征整正

tsik 織職

ɨŋ

3

iᴇŋ

tsiak 勺 妁繳酌灼 tsɨŋ 證烝蒸拯

iak

tsiaŋ 掌嶂章獐障 tsʰiaŋ 彰

3

it

w

tsiun 準稕 tun 肫xxii

iaŋ

in

w

3

3

3

開 合

開

藥

職

開

合

陽

術

諄

Table 3.9 (Continued)

tsʰiuŋ 銃 茺充

tsʰiək 尺斥蚇 tsiək 赤

tsʰɨŋ 秤 tsʰiŋ 稱

tsʰiaŋ 廠 唱菖 娼昌 敞倡 tsiaŋ 𩌬 tsiak 綽

tsiun 蠢 tsʰiun 春 siun 踳 tsʰiul 出

siək 石 鉐碩

siəŋ 誠 城盛筬成

sik 殖植寔

tsiak 杓xxvii siak 芍 sɨŋ 承

siaŋ 嘗裳上 償常尙

siun醇蓴鶉 純

siək 螫釋 tiək 適xxxi

sik 飾 式拭識 siəŋ 聖聲

sɨŋ 勝滕升

siak 鑠

siaŋ 觴晌賞商傷 tiaŋ 塲 hiaŋ 餉xxvi

siun 蕣瞬

sɨŋ 乘繩 iŋ 剩 sik 食蝕

ziuŋ 戎

ziak 箬 iak 弱xxviii若 iŋ 仍

siun 楯 ziun 閏 盾唇 iun 潤 順 siul 術述 tʰiul 秫xxiii ziaŋ 攘壤讓 iaŋ 禳蘘瓤

liuŋ 融xxxv

ʌik 液腋䘸掖 hiək 弈xxxii奕xxxiii iək 易繹驛亦 tʰiək 蜴xxxiv iək 役疫䓈

piŋ 梬xxx iəŋ 贏楹盈嬴 iəŋ 頴營穎

iŋ 孕媵賸 sɨŋ 蠅xxix ik 翼弋翌杙

iak 籥鑰躍爚藥

hiul 鷸xxiv 繘xxv iaŋ 養羊陽洋楊烊 揚恙樣

iun 尹允

合

合

鍾

燭

juk

jṷo̝ŋ

jṷo̝k

3

3

3

[ts]:103 [tsh]:17 [t]:7 [th]:1 [s]:2

tsʰiok 燭

tsiuk 粥 tsʰiuk 祝 tsioŋ 種腫踵鐘鍾

[tsh]:44 [ts]:6 [s]:3 [tʰ]:1

tsʰiok 觸

siuk 俶xxxvi tsʰiuŋ 衝

[s]:88 [ts]:1 [tsh]:3 [k]:1 [th]:1

siok 屬

siuk 熟孰 淑塾

[s]:85 [ts]:3 [t]:2 [ø]:2 [h]:1

sioŋ 舂 tsoaŋ 𪄻xxxvii sok 束

siuk 叔菽

[s]:21 [t]:1 [th]:1 [ø]:1

siok 贖 ziok 褥 iok 辱 HMJH [z]:46 [s]:1 [n]:2 [ø]:6 YH [ø]:22 CJM [ø]:1

ziuk 肉

[ø]:163 [h]:4 [z]:1 [s]:1 [p]:1 [th]:2 [tsh]:1 [l]:1

ioŋ 勇涌蛹踴埇容 傭鎔墉庸踊用 iok 浴谷欲

iuk 育鬻毓

i The SK reading of she 射 is [s*ia] in HMJH (YS and DD edition), and [sia] in CJM, YH, NE, SE, SG, JY and SH. Synn (2006: 65) notes that the tense consonant [s*] in [s*ia] is derived from the artificial representation of the chuan 船 initial, which has the feature [+voice] in MC. ii HMJH [iəi], HD [siəi] (popular reading [iəi]), SU, JU [siəi], OP [siəi] (popular reading [iəi]) iii HMJH, YH, SH [i], HD, OP [si] (popular reading [i]), SU, JU [si] iv HMJH [tiəi], HD, SU, JU [si] v HMJH [iəi], HD [zui] (popular reading [iəi]) vi HMJH, SE, SG [ki], HD, OP [si] (popular reading [ki]), SU, JU [si] vii HMJH, SG [sio] HD, OP [tsio] (popular reading [sio]), SU, JU [tsio] viii YH, SE, SG, JY [sio], HD [sio] ix YH, CJM, SE [thio] x HMJH [‍tʰio‍], SG, HD, SU, JU [tshio] xi HMJH, SG [io] xii HMJH [tsʰiuk], HD, SU, JU [tsu] xiii HMJH, HD [tsiəm], JU [siəm] xiv HMJH (YS edition) [ziəm], HMJH (DD edition) [siəm] xv HMJH, YH [tsʰiəm], HD, OP [iəm] (popular reading [tsʰiəm]), SU, JU [iəm] xvi HMJH, HD [sim], OP[tsʰim] (standard reading [sim]), SU, JU [tsʰim] xvii YH, SE [thim], SG, HD [tshim], OP [sim] (standard reading [tshim]) xviii HMJH [nim], HD [zim], SU, JU [im] xix HMJH [siəl], HD [ziəl] (popular reading [siəl]), SU, JU, OP [siəl] xx HMJH, YH [til], HD, SU, JU [tsil] xxi HMJH, IG [til], HD, SU, JU [tsil]

Summary

合

屋

xxii xxiii xxiv xxv xxvi xxvii xxviii xxix xxx xxxi xxxii xxxiii xxxiv xxxv xxxvi xxxvii

HMJH [tun], JY [tsiun] HMJH [tʰiul], SU, JU [siul], HD, OP [siul] (popular reading [tsʰiul]) HMJH, HD [hiul], OP [iul] (popular reading [hiul]), SU, JU [iul] HMJH [hiul], SU, JU [iul] HMJH [hiaŋ], HD [siaŋ] (popular reading [hiaŋ]), YH, SE [hiaŋ], OP [siaŋ] (popular reading [hiaŋ]) HMJH (YS edition) [tsiak], HMJH (DD edition) [siak], HD [tsiak] HMJH (YS edition) [iak], HMJH (DD edition) [ziak] HMJH, YH, SG[sɨŋ], HD, OP [ɨŋ] (popular reading [sɨŋ]), SU, JU [ɨŋ] HMJH [piŋ], HD [iəŋ], SU, JU [iŋ], OP [iŋ] (standard reading [iəŋ]) YH, CJM, NE, SE, SG [tiək], HD [tsiək], OP [siək] (popular reading [tsiək]), SU, JU [siək] HMJH, YH, HD [hiək], OP [iək] (standard reading [hiək]), SU, JU [iək] YH, SG, HD [hiək], SU, JU [iək] HMJH [thiək], HD, OP [iək] (popular reading [thiək]), SG [siək], SU, JU [iək] YH [liuŋ], HD [iuŋ] (popular reading[liuŋ]) YH, SE, SG [siuk] HMJH [tsoaŋ], SU, JU [sioŋ], HD, OP [sioŋ] (popular reading [tsoaŋ])

Initial system 103 The chan 禪 initial is represented as [s] in EMSK. The following representations are derived from analogy: 鰭 (群 [g] initial) > 嗜 [ki], 沈 (澄 [ɖ] initial]) > 忱 [tʰim], 喘 (昌 [tɕʰ] initial) > 腨 [tsʰiən], 酌 (章 [tɕ] initial) > 杓 [tsiak]. The reading of 擅 [tsʰiən] may derived from the reading of 𧾍, one of the fanqie for which is 直然 (Pak 1971: 84). The shu 書 initial is represented as [s] in EMSK. The exceptional representations that result from analogy are listed as follows: 銳 (以 [j] initial) > 蛻 [iəi], 占 (章 [tɕ] initial) > 痁 [tsiəm], 饘氈 (章 [tɕ] initial) > 羶 [tsiən], 場 (澄 [ɖ] initial]) > 塲 [tiaŋ], 向 (曉 [h] initial) > 餉 [hiaŋ]. There are three fanqie spellings for 適 [tiək] in Guangyun, which are shizhi 施隻, zhishi 之石, duli 都歷. The reading of 適 [tiək] is consistent with the third fanqie spelling. The reading of chi 弛 in EMSK is [i]. There are three fanqie spellings for chi 弛 in Jiyun, namely shangshi 賞是, chouzhi 丑豸 and yuzhi 余支. The reading of 弛 [i] is consistent with the yuzhi 余支 fanqie spelling. The chuan 船 initial is represented as [s] in EMSK. The reading of 䑛 [tiəi] and 秫 [tʰiul] are derived from analogy: 抵低 (端 [d] initial) > 䑛 [tiəi], 𦬸 (澄 [ɖ] initial) > 秫 [tʰiul]. There are two fanqie spellings for 賸, a variant of 剩, in Guangyun, which are yizheng 以證 and shizheng 實證. The reading of 剩 [iŋ] is consistent with the fanqie spelling yizheng 以證 (Kōno 1979: 395). The ri 日 initial is generally represented as [z] in HMJH. The exceptional representations that derived from analogy are listed as follows: 亵褻 (心 [s] initial) > 𤑔 [siəl]. The representation of nen 恁 is [nim] in HMJH. There are four fanqie spellings for nen 恁 in Jiyun, namely renshen 忍甚, ruzhen 如鴆, rulin 如林 and nixin 尼心. The reading [nim] is consistent with the nixin 尼心 fanqie spelling. It should be noted that the ri 日 initial is represented as a zero-initial in YH and CJM (SB edition). In CJM (GJ edition), the ri initial is transcribed as a zero-initial, except in the case of 日 [zil]. The yi 以 initial is represented as a zero initial in EMSK. The exceptional representations that derived from analogy include: 滯 (澄 [ɖ] initial) > 靾 [thiəi], 爓 (邪 [z] initial) > 𢸴 [ziəm], 幨襜 (昌 [tɕʰ] initial) > 簷 [tsʰiəm], 瞲獝 (曉 [h] initial) > 鷸繘 [hiul], 繩 (船 [ʑ] initial) > 蠅 [sɨŋ], 聘 (滂 [ph] initial) > 梬 [piŋ], 赫 (曉 [h] initial) > 奕 [hiək], 剔 (透 [th] initial) > 蜴 [tʰiək] (Pak 1971: 50–52). No consensus has been achieved concerning the reasons for [liuŋ] reading of 融. According to Kōno (1979: 363), the reading of 融 [liuŋ] is considered as a result of false regression of Dueum beopchik [Rule of initials], which regulates that a word-initial /l/ should be dropped when it followed by /i/. In other words, the /l/ initial in 融 [liuŋ] may be a result of hyper correction. Nam (1973: 102–104) argues that the [l] initial of 融 [liuŋ] can be attributed to the assimilation of the [l] coda of the preceding syllable in the notation. In YH, before the notation [liuŋ] is the semantic notation “no kʌl,” which means “melt.” Ito (2007: 106) explains that the yi 以 initial, which is an approximant [j] in the source input, may became [ɾ] in actual speech, and thus the [l] initial of 融 [liuŋ] could be a faithful representation of the [ɾ] in the source input.

104  Initial system

3.10 Summary Based on the above discussion, the mainstream representation of EMSK initials can be summarized as below. Table 3.10  Mainstream representation of EMSK initials

Group 幫 (Bilabials)

Group 非 (Labial-dentals)

Group 見 (Velars)

Group 影 (Gutturals)

Group 端 (Alveolar stops, lateral, and nasal)

Group 精 (Alveolar fricatives, and affricates)

Group 知 (Retroflex stops, and nasals) Group 莊 (Retroflex fricatives, and affricates) Group 章 (Palatal fricatives and affricates)

Initial

Qieyun

LMC

EMSK

幫 滂 並 明 非 敷 奉 微 見 溪 群 疑 影 曉 匣 云 端 透 定 泥 來 精 清 從 心 邪 知 徹 澄 娘 莊 初 崇 生 章 昌 禪 書 船 日 以

p pʰ b m p pʰ b m k kʰ g ŋ ʔ h ɦ ɦ t tʰ d n l ʦ ʦʰ ʣ s z ʈ ʈʰ ɖ ɳ tʂ tʂʰ dʐ ʂ tɕ tɕʰ dʑ ɕ ʑ ȵ j

p pʰ b m f fʰ v ɱ k kʰ g ŋ ʔ h ɦ ø t tʰ d n l ʦ ʦʰ ʣ s z ʈ ʈʰ ɖ ɳ tʂ tʂʰ dʐ ʂ tɕ tɕʰ dʑ ɕ ʑ ʑ ø

p, pʰ p, pʰ p, pʰ m p, pʰ p, pʰ p, pʰ m k k k ø ø h h ø t, tʰ tʰ, t t, tʰ n l ʦ, ʦʰ ʦʰ, ʦ ʦ, ʦʰ s s t, tʰ tʰ, t t, tʰ n ʦ, ʦʰ ʦʰ s, ts, ʦʰ s ʦ, ʦʰ ʦʰ s s s z ø

As can be seen from Table 3.10, due to the mismatch in the phonemic inventory between Chinese and Korean, some initials distinguished in MC are represented as the same type in SK. For example, the MC yi 疑, ying 影, yun 云 initial and the

Initial system 105 yi 以 initial are represented as a zero-initial in SK. The zhuang 莊 series (retroflex affricates and fricatives), the zhang 章 series (palatal affricates and fricatives), and the jing 精 series (alveolar affricates and fricatives) are also represented as one type, namely, alveolar affricates and fricatives. The voiced initials in MC are generally represented as voiceless in SK. However, it should be noted that the ri 日 initial is represented as voiced [z] in some data in MSK. This raises the question that why Old Korean adapted the chuan 船 initial as [s] if it was able to adapt the [ʑ] of the ri 日 initial as [z]. According to Kōno (1979), the representation of the ri 日 initial as [z] is only seen in HMJH and HG among those materials collected before the Imjin war (1592). The ri 日 initial was represented as [ø] in YH, CJM, SU, JU, OP. Pak (1971: 86) argues that as the contrast between voiced and voiceless did not exist in Korean phonology, the [z] representation in HMJH is an artificial convention. The [z] sound in Middle Korean occurs when a [s] is in intervocalic position and thus should not be viewed as a phoneme. Lee (1997: 133) argues that the ri 日 initial [ȵʑ], which is less sonorant than fricative [ʑ] because of its nasal feature, shares the same place of articulation with [j]. It is thus probable that the ri 日 initial [ȵʑ] was adapted in Korean as [j] first, and later represented as [z] artificially. This study agrees to Pak and Lee’s analysis: The ri 日 initial [ȵʑ] was originally adapted as [j], and the [z] representation in HMJH and HG is an artificial practice. One intriguing question about the initial system in EMSK is the aspiration assignment in that it is not consistent with MC. By examining the data in HMJH, we find that none of the following factors, including rhyme, roundness, division or tone in Chinese, is a contributing criterion for aspiration assignment in EMSK. Table 3.11 shows that two characters with exactly the same phonetic form in MC are represented in HMJH differently in terms of aspiration. Therefore, the assignment of aspiration is not related to the source form in MC. Table 3.11  The assignment of aspiration in HMJH Initial

Character

Rhyme group

Rhyme

Kaihe

Division

Tone

Qieyun

HMJH

定 定 定 定 定 定 定 定 精 精 精 精

臺 苔 怠 袋 娣 髢 頭 骰 睫 椄 脊 跡

蟹 蟹 蟹 蟹 蟹 蟹 流 流 咸 咸 梗 梗

咍 咍 咍 咍 齊 齊 侯 侯 葉 葉 昔 昔

kaikou kaikou kaikou kaikou kaikou kaikou kaikou kaikou kaikou kaikou kaikou kaikou

1 1 1 1 4 4 1 1 3 3 3 3

ping ping shang shang qu qu ping ping ru ru ru ru

dəi dəi dəi dəi dei dei du du tsɣiᴇp tsɣiᴇp tsiᴇk tsiᴇk

tʌi tʰʌi tʰʌi tʌi tiəi tʰiəi tu tʰu tsʰiəp tsiəp tsʰiək tsiək

The following table suggests that the assignment of aspiration is related to an analogy based on phonetic components. Characters that share the same phonetic component tend to be assigned to the same value of aspiration.

106  Initial system Table 3.12  Analogy and assignment of aspiration in HMJH Phonetic component

HMJH

The initials of these characters in Qieyun

皮

pʰa 簸波菠, pʰi 彼陂 pʰa 坡, pʰi 鈹 pʰa 蔢婆櫇, phi 被鞁髲疲皮 pʰo 飽苞 pʰo 砲‍‍泡 pʰo 皰‍庖‍匏‍咆‍炮‍跑‍麅‍齙鉋‍ pən 藩 pən 潘旙轓 pan 燔

p pʰ b p pʰ b p ph b

包 番

It is worth noting that in HMJH, there is no single character with a reading of [pa]. If we examine the syllable type of [p] and [ph] initials in HMJH, we can find that there exists preference for specific rhyme types with the initials [p] and [ph]. Table 3.13  Syllable preference and the assignment of aspiration in HMJH

p ph p ph p ph

a

ai

al

an

aŋ

ak

ən

u

un

um

－ pʰa ɨi pɨi － i pi pʰi

－ pʰai ɨk pɨk － io － pʰio

pal pʰal ɨŋ pɨŋ － il － pʰil

pan pʰan o po pʰo iəi － pʰiəi

paŋ － oŋ poŋ － iən piən pʰiən

pak － ok pok － iəŋ － pʰiəŋ

pən － ʌi pʌi － iək piək －

pu pʰu ʌl pʌl － iəl piəl －

pun － ʌiŋ

－ pʰum ʌik pʌik － － － －

pʰʌiŋ ip pʰip －

[p] and [ph] only appear before the rhyme in the following six types: [al], [an], [u], [o], [i] and [iən]. For the other syllable types, only one form, either [p] or [ph], is allowed. This shows that most rhymes have a pattern of co-occurrence with, or preference for [p] or [ph] in HMJH. It is unreasonable to claim that syllable type [pa] did not exist in the phonotactic inventory in Middle Korean. Thus, the gap in these syllable types suggests simplification in the assignment of the aspiration feature. Kōno (1979) proposes that this simplification results from a pursuit of convenience of pronunciation memorization. This is a possible explanation; however, this suggestion does not give explanation to what causes a particular syllable to tend to be represented as aspirated or unaspirated. Since the aim of this study is to examine the origin of SK, we are particularly interested in those features in SK which can give us a hint regarding its origin. The following features in Middle SK, that is, a lack of distinction between groups bang 幫 (bilabials) and fei 非 (labiodentals), groups zhi 知 (retroflex stops and nasal) and duan 端 (alveolar stops and nasal), the yun 云 initial being represented as a zero-initial, can give us insight into the origin of SK. We will provide a detailed discussion on these features in Chapter 6.

4 Final system of Middle Sino-Korean

This chapter will compare the rhyme system in MC and that in EMSK. The rhyme system of MC will be introduced based on the structure found in Qieyun, and the rhyme system of EMSK will be introduced following the structure given in HMJM, CJM and YH. In order to provide a more comprehensive understanding of the rhyme system in MSK, for those non-mainstream representations, the phonetic notations in other materials will also be given in the footnote for reference. The rhymes of Middle Chinese are always gathered into 16 she 摄 (rhyming group), namely, 果假遇蟹止效流咸深山臻宕江曾梗通, according to principal vowel and coda. In this study, the discussion will follow the order of these 16 rhyming groups. In each comparison table, we will provide the information about kaihe “labial roundness” and deng “division” in the first line, the reconstructed phonetic value introduced in Pan and Zhang (2013) in the second line, and the representation in EMSK in the third line. It should be noted that besides the normal first, second, third and fourth divisions, there are also the divisions of 3A and 3B. The divisions 3A and 3B are used to indicate the chongniu “rime doublets” in MC, which refers to the fact that there are two relatively different fanqie spellings in Qieyun under certain rhymes. Chongniu only occur in the labial, velar or glottal initial of the following eight rhymes: zhi 支, zhi 脂, xiao 宵, ji 祭, xian 仙, yan 鹽, zhen 真 and qin 侵. In the rhyme tables, they are placed in the third and fourth division, respectively. Traditionally, those which are placed in the fourth division are called 3A division, and those which are placed in the third division are called 3B division.

4.1  Guo 果 rhyme group Except for the case of 左佐 [tsoa], the ge 歌 rhyme (kaikou div. 1) is fully represented as [a] in EMSK. The reading of 左佐 is [sa] in Kan-on and Go-on. Kōno (1979: 427) states that 左 [tsoa] may be a reflection of the sound change [tsa] > [tsuə] in the transition period from Middle Chinese to Old Mandarin. Lee (1980: 125) claimed that the [tsoa] reading is derived from the influence of Old Mandarin. However, the reading of [tsa] underwent the following change from Middle Chinese to Modern Chinese: [tsa] > [tsɔ] > [tso] > [tsuo] > [tsuə] (Wang 1985). If the [o] in [tsoa] is from the influence of Old Mandarin, then the adapted form in HMJH should had been [tso], a reading that we can find in NGD and PTS. Ito

108  Final system Table 4.1 The guo 果 rhyme group in EMSK kaikoui div. 1 歌

hekou div. 1 戈

hekou div. 3 戈

Qieyun

ɑ

ʷɑ

ʷiɑ

EMSK

a

幫 滂 並 明 見

p pʰ b m k

溪 群 疑 影 曉 匣

kʰ g ŋ ʔ h ɦ

端 透 定

t tʰ d

泥 來 精

n l ʦ

清 從 心

ʦʰ ʣ s

oa

ka 珂可軻

tsʰa 搓磋 tsa 鹺咱vi sa 挱

o, ai, oai

pʰa 簸波菠播 pʰa 坡破頗 pʰa 蔢婆櫇 ma 魔磨蘑摩 koa 菓戈堝腡鍋 果裹過 koa 課騾科蝌窠

ka 歌柯哥

a 我餓蛾鵞 a 阿 ha ha 河荷魺 何苛賀 ta 多 tʰa 他拖 ta 爹 tʰa 駝舵駄 跎鼉 na 挪 la 羅蘿籮鑼

oa, a

ka 瘸

oa 臥訛 oa 渦倭窩萵 hoa 火貨夥 hoa 禾和禍 tʰa 𦖋 tʰa 涶唾 tʰa 惰 墮

tsoa 左iv佐v

oa, a

hoa 靴

tʰai 妥ii

na 懦糯捼 la 裸蓏瘰螺𥡜

lo 騾iii

tsoa 坐矬 tsʰa 銼 soa 鎖vii sa 梭viii蓑ix

soai 瑣x

i In this study, the description on the kaikou and hekou follows the book Fangyan diaocha zibiao 方言調查字表 [Questionnaire of characters for dialect survey]. ii YH [tʰai], SG [tʰa] iii YH, HD, SU, JU, CJM (SB edition) [la], HMJH, CJM (GJ edition) [lo] iv  HMJH, YJ, SH, YH, SE, IG, CJM (GJ edition) [tsoa], DH, NE, SG, JY, HD, SU, JU, CJM (SB edition) [tsa], OP [tsa] (popular reading [tsoa]) v SH, YH, CJM (GJ edition) [tsoa], SG, CJM (SB edition) [tsa] vi In addition to [tsa], HMJH also lists the notation [tsam]. vii HMJH [soa], YH [soai] viii HMJH, YH [sa] ix HMJH, SG [sa] x YH, IG, SG [soai]

(2007) explained the origin of 左 [tsoa] from the perspective of coarticulation, arguing that the [o] in 左 [tsoa] results from the [+rounded] feature of the back vowel [ɑ]. We think that Ito’s (2007) analysis makes more sense. Since both readings [tsoa] and [tsa] can be found in MSK data, the [tsoa] reading was probably a variant first, and later became the dominant reading. The ge 戈 [ʷɑ] rhyme (hekou div. 1) is represented as [a] or [oa] generally according to the articulation place of the initials. For labial initials, the [ʷɑ] is represented as [a]. This is consistent with the fact that there is no contrast of roundness in the

Final system 109 labial position in the Qieyun, in which both kaikou initial characters and hekou initial characters can serve as the lower character of fanqie for labial initial characters. For velars and gutturals, the [ʷɑ] is represented as [oa]. For alveolar stops, laterals and nasals, div. 1 戈 [ʷɑ] is represented as [a]. This may be explained by the phonotactic constraint in Korean that does not allow syllable types like [toa], [tʰoa], [noa] and [loa]. For alveolar fricative and affricate initials, both [oa] and [a] can be found. The reading of 騾 is [la] in YH, HD, SU, JU, CJM (SB edition) and [lo] in HMJH and CJM (GJ edition), a reading also seen in NGD and PTS, indicating that the [lo] reading is probably a reflection of the pronunciation in Old Mandarin. The extra [i] in 妥 [tʰai] and 瑣鎖 [soai] may arises from the influence of the following grammatical word, such as the nominative case marker -i and the predicative case marker -i(ra) (Kōno 1979: 447). The ge 戈 rhyme (hekou div. 3) is represented as [oa]. The reading of 瘸 [ka] probably results from the analogy of 加 [ka] (Pak 1971: 121). Table 4.2 summarizes the mainstream representation of the guo 果 rhyme group in EMSK. Table 4.2  Mainstream representation of the guo 果 rhyme group in EMSK

Qieyun EMSK

kaikou div. 1 歌

hekou div. 1 戈

hekou div. 3 戈

ɑ a

ʷɑ Other oa

ʷiɑ oa

Group 幫端 a

4.2  Jia 假 rhyme group Table 4.3 The jia 假 rhyme group in EMSK kaikou div. 2 麻

hekou div. 2 麻

kaikou div. 3 麻

Qieyun

ɣ

æ

ʷɣæ

iæ

EMSK

a, ai, ia

oa, oai

ia, iə, a

幫

p

滂 並 明 見

pʰ b m k

溪 疑 影 曉 匣

kʰ ŋ ʔ h ɦ

pʰa 弝欛靶芭笆疤把 pʰai 壩霸 pʰa 帕𧵅葩舥怕 pʰa 琶耙爬罷杷 ma 瑪螞馬罵麻蟆媽 ka 嫁架稼駕斝椵假 家袈茄枷痂笳葭耞 嘉加

知 徹 澄

ʈ ʈʰ ɖ

a 砑牙芽衙雅迓訝 a 瘂鴉 ha 罅 hia嚇i ha 下廈夏遐瑕蝦霞鰕 ka 暇 tʰa 打 tʰa 詫 ta 茶 tsʰa 搽

koa 寡苽蝸瓜 koa 銙胯誇 oa 瓦𡧗 oa 蛙窪搲 hoa 化樺花 koa 踝 hoa 華鏵 tsoa 檛

(Continued )

110  Final system Table 4.3 (Continued) kaikou div. 2 麻

hekou div. 2 麻

kaikou div. 3 麻

Qieyun

ɣ

æ

ʷɣæ

iæ

EMSK

a, ai, ia

oa, oai

ia, iə, a

精

ʦ

從 心 邪 莊

ʣ s z tʂ

初 崇 生 章

tʂʰ dʐ ʂ tɕ

昌 禪 書 船 以

tɕʰ dʑ ɕ ʑ j

tsa 鮓蚱炸榨皶 sa 詐渣 tsʰa 齇𤹡 tsia 樝 tsʰa 岔汊靫杈叉差 sa 槎乍查 sa 沙裟紗鯊

tsʰia 借且 tsiə 姐罝 tsʰa 嗟 tsia 褯 tsʰia 藉 sia  寫瀉卸 sia 斜邪謝

soa 耍

tsia 赭柘炙者 tsʰia 遮 tsʰia 車 sia 社 sia 舍賒奢捨赦 sia 射麝蛇 ia 夜㖡野冶爺也

i  YH [hia], SU, JU, OP [ha]

The ma 麻 rhyme (kaikou div. 2) and the ma 麻 rhyme (hekou div. 2) are generally represented as [a] and [oa], respectively. The reading of 霸 is [phai] in NE, CJM and OP. The reading of 壩 is [pʰai] in HMJH, and [pʰa] in SU, JU and HD. Lee (1997: 141) and Shin (2015: 83) propose that the reading of 霸 壩 [pʰai] could be a borrowing from Old Chinese, in which some yu 魚 group characters merged with ge 歌 group characters. Since the reading of the character 欛 is [pʰa] in HMJH, the question arises why only 壩霸 conserve the Old Chinese reading. As suggested by Ito (2007: 117), the reading of 壩霸 [pʰai] can be better explained by hypercorrection. The reading of 嚇 is [hia] in YH, but [ha] in SU, JU, OP. The [hia] reading in YH is a reflection of Old Mandarin, as seen in the reading [xia] in ZYYY (Yang 1981: 158; Ning 1985: 113). Zha 樝 is represented as [tsia] in the YS edition of HMJH, but as [tsa] in SU, JU and the JGG edition of HMJH. The [tsia] reading of 樝 is probably a reflection of the reading [tʂa] in Old Mandarin and the [i] glide in [tsia] is inserted to mimic the retroflex initial [tʂ] (Shin 2015: 90). The ma 麻 rhyme (kaikou div. 3) is generally represented as [ia]. The reading of 姐罝 in MSK is given as follows: 姐 SU, JU, HD [tsia], HMJH [tsiə]; 罝 SU, JU, HD [tshia], HMJH [tsiə]. The reading of 姐罝 [tsiə] in HMJH could be a result of the analogy of 且 [tsiə] (Lee 1997: 142; Ito 2007: 118). The reading of 嗟 is [tsʰa] in YJ, YH, HD, IG, SH and [tsʰia] in SG. Both [tsʰa] and [tsʰia] can be found for 嗟 in SE. The [tsʰa] reading is a reflection of the dropping of -j glide in Middle Korean. Table 4.4 is a summary of the mainstream representation of the jia 假 rhyme group in EMSK.

Final system 111 Table 4.4  Mainstream representation of the jia 假 rhyme group in EMSK

Qieyun EMSK

kaikou div. 2 麻

hekou div. 2 麻

kaikou div. 3 麻

a a

ʷɣa oa

ia ia

ɣ

4.3  Yu 遇 rhyme group 4.3.1 Mu 模 rhyme Table 4.5 The mu 模 rhyme in EMSK hekou div. 1 Qieyun

ṷo

EMSK

o

幫 滂 並 明 見

p pʰ b m k

溪 疑 影 曉 匣

kʰ ŋ ʔ h ɦ

端 透 定 泥 來

t tʰ d n l

精 清 從 心

ʦ ʦʰ ʣ s

po 圃逋補 pʰo 布晡 po 浦普 pʰo 怖 po 捕步𩊶捗 pʰo 哺酺蒲葡蒱 mo 姥暮模氁慕謨 ko 股賈瞽罟鼓蛊牯羖顧雇錮孤沽 辜姑箍蛄古𦍩菇故固蠱痼 ko 苦庫袴枯 o 五伍寤捂䎸吾蜈鼯杇誤悟忤迕 o 烏洿𦶀汙 ho 琥戽虎呼 ho 户帍簄枑狐瑚壺猢湖弧糊箶衚 鬍鶘瓠蝴護互胡冱乎 to 賭覩都 tʰo 土吐菟兔 to 鍍度屠徒途圖杜渡塗 no 弩奴 lo駑怒 lo 櫓虜滷輅鷺路露艫蘆轤顱鱸纑 鸕攎爐𤬛 no 艣 tso 祖租組 tsʰo 醋 tso 措 tso 祚粗 so 訴嗉素塑蘇酥櫯遡泝

i ii iii iv v vi

u pu 溥i pʰu 鋪ii pu 簿iii埠

io, oi

mio 墓

io 梧

tu 蠹iv tʰu 妬v tu 肚vi loi 賂

tsʰu 麤

YH, HD [pu], JY [po], SG [po]/[pu] HMJH, HD [pʰu], YH, SU, JU [pho] HMJH, SH, SU, JU, HD [pu] HMJH, YH, SE, SG [tu] YH, YJ, SH [tʰu] HMJH, SG [tu]

The mu 模 rhyme is generally represented as [o] except in the following characters: 溥 [pu], 簿 [pu], 埠 [pu], 肚 [tu], 鋪 [pʰu], 蠹 [tu], 妬 [tʰu] and 麤 [tsʰu]. The reading for the mu 模 rhyme is [u] in NGD and PTS, such as 布步哺 [pu], 蒲 鋪 [pʰu], 肚 [tu]. The [u] representation is regarded as a reflection of the Chinese

Final system

112

pronunciation in Old Mandarin in Lee (1980: 129) and Lee (1997: 249). However, the shift from [o] to [u] for the mu 模 rhyme started from the Tang Dynasty (Wang 1985), therefore, the [u] reading by itself does not necessarily mean it is a reflection of Old Mandarin. In EMSK, the hun 魂 rhyme has both [on] and [un] readings, and the mo 沒 rhyme has both [ol] and [ul] readings. The [u] reading in the mu 模 rhyme should be in parallel relationship to the [un] reading in the hun 魂 rhyme and the [ul] reading in the mo 沒. The [u]/[un] reading generally occurs after labial and alveolar initials in the mu 模 rhyme and the hun 魂 rhyme, implying that it is probably a result of misperception in the process of adaptation since the [ṷ] in [ṷo] is relatively more salient in these phonological environments. The reading of 梧 is [o] in HMJH (DD, JGG, NG, GJG, DG edition), YH, CJM, HD, SU, JU. The [i] of [io] 梧 in the YS edition of HMJH is probably an erroneous inscription. The reading of 墓 is [mio] in YJ, HMJH, SH, YH, SU, JU, SG, SE, and [mu] in HD. The reading of [mio] 墓 “grave, tomb” is probably derived from the reading of 廟 [mio] “ancestral temple,” which is semantically related (Kōno 1979: 491). The reading of 賂 is [lo] in HD, SU, JU and [loi] in HMJH, YH, SG. The [loi] reading may be derived from the analogy of 賄 [hoi], which is semantically close to 賂 [loi] (Ito 2007: 124). 4.3.2 Yu 魚 rhyme Table 4.6 The yu 魚 rhyme in EMSK hekou div. 3 Qieyun

iɤ

EMSK

iə

見

k

溪 群

kʰ g

疑 影 曉 來

ŋ ʔ h l

知 徹 澄 娘 精 清

ʈ ʈʰ ɖ ɳ ʦ ʦʰ

從 心

ʣ s

liə 侶膂旅濾鑢 閭廬驢呂慮 tiə 瀦貯豬 tiə 楮樗摴 tiə 苧杼筯儲 niə 女 tsiə 蛆苴 tsiə 疽沮鴡 tsʰiə 覷 tsiə 咀 siə 絮胥

ə kə 鋸車居踞據 裾擧 kə 袪去 kə 距炬苣 漫渠蕖 𪆂拒巨遽 ə 圄御語馭魚漁 ə 飫於 hə 虛噓許

tʰə 攄

o

əi, iəi, iu, ia kəi 筥i

tiəi 豬ii tiəi 除iii

tsiu 𡲰

Final system 113 hekou div. 3 Qieyun

iɤ

EMSK

iə

邪

z

莊 初 崇

tʂ tʂʰ dʐ

siə 嶼芋序漵叙 緖徐 tsiə 詛

生

ʂ

章

tɕ

tsiə 渚

昌 禪 書

tɕʰ dʑ ɕ

船 日 以

ʑ ȵ j

tsʰiə 處 tsiə 杵 siə 墅署曙薯 siə 鼠暑黍𧑓恕 書鵨庶舒 siə 抒紓 ziə 汝洳 iə 茹如 iə 蕷予蜍鸒輿 餘余與藇

i ii iii iv v vi vii viii ix x

HMJH [kəi], SG, HD, SU, JU [kə] HMJH [tiəi], SH, YH [tiə], SU, JU, HD [tsiə] YH [tiəi], HG [tiə], SH [tiəi]/[tiə] HMJH, YH[sə], SU, JU [tso], HD [tsu] (popular reading [tso]) HMJH [tsia], HD[tsiə] (popular reading [tsia]), SU, JU [tsiə] HMJH [tsia], HD, SU, JU [tsiə] HG, YH, CJM, SE, IG, SH [tsiəi], NE, JY, HD, SU, JU [tsiə], SG [tsiəi]/[tsiə] HMJH, YH, SE, CJM [iəi], IG [iəi]/[iə], SG, JY, HD, SU, JU [iə], OP [iə] (popular reading [iəi]) SH, YH, OP [iəi] SH, JY, YG, CJM, SE, IG, SG [iəi], HD, SU, JU [iə], OP [iə] (popular reading [iəi])

ə

o

sə 鋤iv

tso 葅阻 tsʰo 礎初楚 tso 助 so 所䟽蔬 梳踈

əi, iəi, iu, ia

tsia 煮v翥vi tsiəi 諸vii

iəi 譽viii預ix豫x

The yu 魚 rhyme is generally represented as [ə] when the initial is velar or guttural, [o] when the initial is retroflex fricative or affricate and [iə] elsewhere. The readings that derived from analogy include 者 > 煮翥 [tsia]. The yu 魚 rhyme is mainly represented as [o] when the initial belongs to the zhuang 莊 series. In Pak (1971) and Lee (1997), this [o] is explained as a reflection of the [ï̱ o] in Old Chinese, but as a borrowing of Old Mandarin in Arisaka (1936) and Kōno (1979). However, this [o] may not derive from different layers of borrowing, but could be a result of adjustment in the process of adaptation. Due to the influence of the zhuang initials, which are retroflexes, the main vowel of yu 鱼 rhyme become more back and rounded, and thus borrowers tend to use [o], which has [+back] and [+round] feature, to transcribe this sound. The reading of 鋤 [sə] is a reflection of the dropping of glide [j] after [ts], [tsh] and [s] in Middle Korean. The insertion of [i] ending in 豬除 [tiəi], 諸 [tsiəi] and 譽預豫 [iəi] may arises from the assimilation of the medial [i]. 𡲰 [tsiu] is borrowed from Old Mandarin. The final [iu] in [tsiu] is a reflection of the final part [y] in Old Mandarin (Kōno 1979: 493; Ito 2007: 127).

114  Final system 4.3.3 Yu 虞 rhyme Table 4.7 The yu 虞 rhyme in EMSK hekou div. 3 Qieyun

iṷo̝

EMSK

u

非 敷

f fʰ

奉 微

v ɱ

見 溪 群 疑 影 曉 云 來 知 澄 精 清 心 從 初 崇 生 章 昌 禪 書 日 以

k kʰ g ŋ ʔ h ɦ l ʈ ɖ ʦ ʦʰ s ʣ tʂʰ dʐ ʂ tɕ tɕʰ dʑ ɕ ȵ j

pu 府俯斧賦傅夫膚跗柎付 pu 俘郛麩稃莩筟敷赴 mu 撫 pu 父腐釜鮒駙符芙鳧扶附 mu 鵡武舞廡務霧巫誣無 騖蕪 ku 句駒矩拘俱 ku 軀驅區 ku 衢癯臞蠷具懼窶 u 愚虞隅寓遇 u 紆迂 ku 嫗 hu 酗煦 u 羽雨宇芋盂栩禹 lu 縷褸縷屢 tsu 註

i ii iii iv v vi vii viii

tsʰu 趨 tsʰu 芻 tsʰu 雛 su 數 tsu 炷注鑄

iu

iui, o pʰo 脯i po 輔ii

kiu 𣂐iv

mo 侮iii

o 娛v

tiu 拄株蛛誅 tiu 柱幮廚住 siu 鬚須

tsʰiui 取vi 娶vii tsʰiui 聚viii

tsiu 主蛀朱珠 tsʰiu 樞 siu 樹銖殊豎 siu 輸 ziu 乳孺儒 iu 襦濡 iu 諭諛榆逾庾愈愉踰裕

HMJH, SG, NE [pho], SH [pho]/[po] SH, NE, YH, SE, IG, SG, SU, JU [po], HD [pu] YH, HG, SE, SG, NE, HG [mo], HD [mu] HMJH [kiu], SU, JU [ku] SH, YH, SG [o], OP [u] (popular reading [o]) YJ, SH, DH, JG, YH, NE, SE, IG, SG, JY [tshiui], CJM [tsiui], HG, HD, SU, JU [tshiu] It is written as 娵 in HMJH. HMJH, SH, YH, SE [tshiui], HD, SU, JU [tshiu] YG, SH, DH, NE, YH, CJM, IG [tshiui]

The yu 虞 rhyme is generally represented as [u] when the initial is labial, velar, guttural, lateral or a retroflex fricative/affricate and [iu] elsewhere. The [u] reading in 炷注鑄 [tsu] is a reflection of the dropping of glide [j] after [ts], [tsh] and [s] in LMK. The readings that derived from wrong analogy include 捕浦 > 輔 [po], 哺晡酺 > 脯 [pʰo], 吳 > 娛 [o]. The reading of 取娶 in MSK is listed as follows: 取 YJ, SH, DH, JG, YH, NE, SE, IG, SG, JY [tshiui], CJM [tsiui], HG, HD, SU, JU

Final system 115 [tshiu]; 娶 SH, YH, SE [tshiui], HD, SU, JU [tshiu]. Although the expected reading of 取娶 is [tsʰiu], it is represented as [tsʰiui] in many MSK materials. There is a similar situation in hekou zhi 支 rhyme character 炊 [tsʰiui], hekou zhi 脂 rhyme character 醉 [tsʰiui] and kaikou you 尤 rhyme character 僦 [tsʰiui], implying that the initial [tsʰ] causes the nuclear vowel being more fronted and rounded, and thus an [i] ending is inserted. 4.3.4 Summary Below is a summary of the representative reading of the yu 遇 rhyme group in EMSK. From this table, we can find that EMSK preserves the distinction between yu 魚 rhyme and yu 虞 rhyme. Table 4.8  Mainstream representation of the yu 遇 rhyme group in EMSK

Qieyun EMSK

模

魚

u̯ o o

iɤ other iə

虞 velar, guttural ə

retroflex fricative/ affricate o

iu̯ o̝ other iu

labial, velar, guttural u

4.4  Xie 蟹 rhyme group 4.4.1 Hai 咍 rhyme Table 4.9 The hai 咍 rhyme in EMSK kaikou div. 1 Qieyun

əi

EMSK

ʌi

ai

見 溪 疑 影 曉 匣 端 透 定

k kʰ ŋ ʔ h ɦ t tʰ d

hʌi 荄該 kʌi 改 hʌi 咳 kʌi 開

kai 概漑槩 kai 鎧慨 ai 礙 ai 埃

泥 來 精 清

n l ʦ ʦʰ

ʌi 愛哀 hʌi 海醢 hʌi 孩頦 tʌi 戴 tʌi 貸 tʰʌi 態胎 tʰʌi 怠苔笞炲殆 tʌi 袋黛代玳擡臺待岱 lʌi 麳來 tsʌi 宰𣪮載栽再災哉 tsʰʌi 綵菜採彩

oi

ɨi

tai 歹

nai 鼐乃耐

loi 賚i sɨi 猜ii (Continued )

116  Final system Table 4.9 (Continued) kaikou div. 1

從 心

Qieyun

əi

EMSK

ʌi

ʣ s

tsʌi 裁財才材纔在 sʌi 塞賽

ai

oi

ɨi sɨi 顋iii

i HMJH, YH, SG, SU, JU [loi], HD [lʌi] (popular reading [loi]) ii YH [sɨi], HD [sai] (popular reading [sɨi]), OP [tshʌi] (popular reading [sɨi]) iii HMJH [sɨi], HD [sai] (popular reading [sɨi]), JU [sʌi]

The hai 咍 rhyme is generally represented as [ʌi]. In the following characters, the hai 咍 rhyme is represented as [ai]: 概鎧漑槩慨 [kai], 礙埃 [ai], 鼐乃耐 [nai] and 歹 [tai]. In HMJH (YS edition) the notation for 埃 [ai] states that “the standard reading of ai 埃 is [ʌi].” The [ai] reading of hai 咍 rhyme is a reflection of the sound change of [ʌ] to [a] in Middle Korean. The reading of 賚 [loi] may derived from the analogy of 賂 [loi], which is close to 賚 [loi] semantically (Kōno 1979: 427). The reading of 顋 [sɨi] may derived from the analogy of zhi 之 rhyme characters 偲緦 [sɨi] (Pak 1971: 138; Ito 2007: 133). Given that 猜 is a zhi 之 group character in Old Chinese, the reading [sɨi] of 猜 may be a reflection of Old Chinese (Lee 1997: 169–170; Shin 2015: 111). 4.4.2 Hui 灰 rhyme Table 4.10 The hui 灰 rhyme in EMSK hekou div. 1 Qieyun

uoi

EMSK

ʌi

幫 滂 並 明

p pʰ b m

溪 疑 影 曉 匣

kʰ ŋ ʔ h ɦ

pʌi 背輩杯 pʌi 坯醅胚配 pʌi 焙培陪倍徘 mʌi 每妹昧苺黴 梅媒煤莓酶

端 透 定 泥 來

t tʰ d n l

tʌi 碓對v nʌi 內vi lʌi 耒viii

oi

ai, oai, ɨi, ui, uəi, iu

pʰai 佩i悖ii koi 傀塊魁 hoi恢 oi 椳偎煨 hoi 賄誨灰晦悔 koi 槐 hoi 茴廻 匯回徊 tʰoi 鎚堆 tʰoi 腿退蓷煺 tʰoi 頹㿉 noi 餒 loi儡擂雷罍磊

ɨi 磑iii ui 桅iv kuəi 潰v

Final system 117 hekou div. 1

清 從 心 i ii iii iv v vi vii viii ix

Qieyun

uoi

EMSK

ʌi

ʦʰ ʣ s

oi

ai, oai, ɨi, ui, uəi, iu

tsʰoi 催淬 tsoi 罪 tsʰoi 摧

siu 焠ix soai 碎

HMJH, YH, SU, JU, SG [p ai], HD [p ʌi] YH, HG, IG, SG, JY, SH [phai], HD [phʌi] HMJH [ɨi], SU, JU, HD [ʌi] HMJH [ui], SU, JU, HD [oi] YH, SG [kuəi], HD, OP [hoi] (popular reading [kuəi]) HMJH, YH, CJM, SU, JU, HD [tʌi] HMJH, YH, CJM, SU, JU, HD [nʌi] HMJH (YS edition) [lʌi], HMJH (DD edition) [loi] HMJH [siu], HD [siui] h

h

The hui 灰 rhyme is generally represented as [ʌi] when the initial is labial, and [oi] elsewhere. There is no kaikou and hekou contrast for labial initials in Middle Chinese. The [ʌi] reading after labial initials should be regarded as a faithful representation of the source language. Both [ʌi] and [oi] can be found when the initial has an alveolar place of articulation. Shin (2015: 85) suggests that [oi] is a layer which corresponds to the Southern Dynasties, while [ʌi] is a layer corresponding to the Tang Dynasty. However, in HLYY, 潰, 對, 會, 悔, 晦, 膾, 碓, 猥, 塊, 碎, 昧, 佩, 貝, 妹, 每, 輩 use the same series of lower characters in fanqie (Huang 1931: 51). In Sino-Vietnamese, the reading of hui 灰 rhyme is generally [oi], such as 對 [doi], 隊 [doi], 雷 [loi], 內 [noi], 罪 [toi], 每 [moi], 塊 [khoi], 媒 [moi] (Wang 1948: 35). The [ʌi] reading is not likely to be a reflection of the pronunciation in the Tang Dynasty. The reading of 碓對 [tʌi] and 內 [nʌi] is probably a result of misperception. These three characters share the qu tone, and this causes the nuclear vowel hard to be perceived. The reading for 耒 [lʌi] is [loi] in YH and the DD edition of HMJH. The reading of 佩悖 [pʰai] is a reflection of the sound change [ʌ] > [a] in Middle Korean. The readings that derived from analogy include 危 (支 rhyme) > 桅 [ui], 粹 (脂 rhyme) > 焠 [siu], 螘 (支 rhyme) > 磑 [ɨi], 貴 (微 rhyme) > 潰[kuəi]. 4.4.3 Tai 泰 rhyme Table 4.11 The tai 泰 rhyme in EMSK

幫 滂 見 疑

kaikou div. 1

hekou div. 1

Qieyun

ɑi

w

EMSK

ai

p pʰ k ŋ

pʰai 貝 pʰai 沛 kai 蓋丐 ai 艾

ʌi

oi, oai

ɑi

oi

hoi 儈膾檜 oi 外 (Continued)

118  Final system Table 4.11 (Continued)

影 匣 端 透 定 泥 來 精

kaikou div. 1

hekou div. 1

Qieyun

ɑi

w

EMSK

ai

ʔ ɦ t tʰ d n l ʦ

ai 藹 hai 害

ʌi

tʰai 汰泰 tʰai 釱 tai大 nai 柰

oi, oai

ɑi

oi hoi 繪會

tʌi 帶i

loi 癩ii瀨iii賴iv tsʰoai 蔡v

loi 酹 tsʰoi 最

i HMJH, YH, CJM, SU, JU [tʌi], HD [tai] ii HMJH [loi], HD [lai] iii HMJH [loi], HD [lai] iv YH, CJM, SE [loi], HD [lai] v YH [tsʰoai], NE, SE [tsʰai]

The tai 泰 rhyme (kaikou div. 1) and the tai 泰 rhyme (hekou div. 1) are represented as [ai] and [oi], respectively. The reading of 帶 [tʌi] is probably an erroneous reading though the reason of this erroneous reading is not clear yet. 帶 [tʌi] was a yue 月 group character in Old Chinese, and it carries [a]/[ɑ] as its nuclear vowel from Old Chinese through to the present day. The reading of 癩瀨 [loi] can be explained by misperception. In other words, the vowel of tai 泰 rhyme is perceived to take the [+round] feature when preceded by the lateral initial lai 來 (Ito 2007: 133). The reading of 蔡 is [tsʰoai] in YH and [tsʰai] in NE and SE. The reason of the insertion of the -o medial in YH is not clear. 4.4.4 Jie 皆 rhyme Table 4.12 The jie 皆 rhyme in EMSK kaikou div. 2 Qieyun

ɣ

EMSK

ʌi

幫 滂 並 明 見

p pʰ b m k

pʌi 拜

溪 疑 影

kʰ ŋ ʔ

hekou div. 2

ɛi

pʌi 排 mʌi 霾埋 kʌi 稭皆 hʌi 痎偕 kʌi 揩

wɣ

ai

pʰai 鞴 mai 邁 kai 疥介芥㠹 ai 騃 ai 噫挨

iəi

ɨi

ɛi

uəi, oi

pɨi 憊i kiəi 界ii階iii犗 誡iv戒v

koi 乖怪 koi 聵vi

Final system 119 kaikou div. 2 Qieyun

ɣ

EMSK

ʌi

匣

ɦ

hʌi 骸諧駭

莊 崇

tʂ dʐ

tsʌi 𢋿齋

i ii iii iv v vi vii viii ix

hekou div. 2

ɛi

wɣ

ai

iəi

ɨi

uəi, oi hoi 懷 koi 壞

hiəi 薤vii

tsiəi 儕viii

ɛi

sɨi 豺ix

HMJH [pɨi], HD, SU, JU [pʌi] HMJH, YH, HD, SU, JU [kiəi] HMJH, YH [kiəi], HD [kʌi] (popular reading [kiəi]), SU, JU [kʌi] YH, CJM, IG [kiəi] YH, SE, IG, SG, JY [kiəi] In addition to [koi], HMJH also lists [huəi]. HD[hoi], SU, JU[oi] HMJH[hiəi], YH, HD, SU, JU[hʌi] HMJH [tsiəi], HD [tsʌi] (popular reading [tsiəi]), SU, JU [tsʌi] HMJH [sɨi], HD [sai] (popular reading [sɨi]), SU, JU [tsʌi]

The jie 皆 rhyme (kaikou div. 2) is generally represented as [ʌi], [ai] or [iəi]. The reprehensive reading of jie 皆 rhyme (hekou div. 2) is [oi]. The [ai] reading for the kaikou jie 皆 rhyme is a reflection of the merging of [ʌ] and [a] in Late Middle Korean. The readings that derived from analogy include 備 (脂 rhyme) > 憊 [pɨi], 齊 (齊 rhyme) > 儕 [tsiəi], 豸 (紙 rhyme) > 豺 [sɨi]. The [iəi] reading for velar initial characters is viewed as a result of the influence of Old Mandarin (Kōno 1979；Lee 1997). In NGD and PKS, the pronunciation of jie 皆 rhyme is [iəi], such as 皆堦街解戒芥疥界 [kiəi], 鞋蟹械薤 [hiəi]. This analysis is acceptable for the explanation of the [iəi] reading (velar initial) of the jie 皆 rhyme. However, if we look at the whole phonological system of SK, we can find the weakness of this explanation. In the geng 庚 rhyme and the geng 耕 rhyme, we can also find the [ʌi]/[iəi] contrast, such as [kʌiŋ] 羹 vs [kiəŋ] 梗綆莄粳鶊, [hʌiŋ] 杏行胻桁 vs [hiəŋ] 衡, [kʌiŋ] 莖 vs [kiəŋ] 耕, [hʌik] 覈核 vs [hiək] 革. Given that the [k] coda has lost in Old Mandarin, the [iək] reading was definitely not borrowed from Old Mandarin. Therefore, the [iəi] reading (velar initial) of the jie 皆 rhyme was not a reflection of Old Mandarin. Another hypothesis is that the [i] in the [iəi]/[iəŋ]/[iək] reading is a result of sound change in Korean. A [j] glide was generated after velar initials. However, in the hai 咍 rhyme, which is generally represented as [ʌi], there is no such [ʌi]/[iəi] contrast. In sum, the [iəi]/ [iəŋ]/[iək] reading is not a result of inner sound change nor a reflection of Old Mandarin. This study explains the origin of [iəi] from the perspective of perception. The [ɣ] glide after velar/guttural initials is more salient and sounds like [j] and therefore was adapted as [i]. 4.4.5 Jia 佳 rhyme The jia 佳 rhyme (kaikou div. 2) and jia 佳 rhyme (hekou div. 2) are represented as [ai] and [oai], respectively. The [iəi] reading in 鞋 [hiəi] is a representation of

120  Final system Table 4.13 The jia 佳 rhyme in EMSK kaikou div. 2

hekou div. 2

ɛ

ʷ ɣɛ

Qieyun

ɣ

EMSK

ai

滂 pʰ 並 b 明 m 見 k

pʰai 派 pʰai 稗𪐄 牌𥱼 mai 賣 kai 街 ai 涯崖 ai 矮

ʌi

iəi

mʌi 買蕒 hʌi 廨懈 解

a, ɨi, ʌik, oai

oai

ka 佳i

koai 拐 枴掛

oi

oa

疑 影 曉 匣 娘 莊 初 崇 生

ŋ ʔ h ɦ ɳ tʂ tʂʰ dʐ ʂ

i ii iii iv v vi vii viii

YH, CJM [ka] HMJH [ʌik], YH, HD, SU, JU [ʌi] HMJH (YS edition) [koi], HMJH (DD edition) [oai] HMJH [hiəi], HD [hai], SU, JU [hʌi] HMJH [hoa], HD, SU, JU [hoai] HMJH [tsʰa], SU, JU, HD [tsʰai] HMJH [sɨi], HD [sai] (popular reading [sɨi]), SU, JU [tsʌi], OP [tsʌi] (popular reading [sɨi]) SG, SH [sai], YH [soai]

nai 嬭 tsʰai 債跐 tsʰai 𨢦

ʌik 隘ii hʌi 蟹邂

hiəi 鞋iv

koi 歪iii

hoa 畫v

tsʰa 釵vi sɨi 柴vii soai 灑viii

the reading in the source language, which is reconstructed as [ɦɣɐi]. The [ɣ] glide after velar/guttural initials is more salient and sounds like [j] and therefore was adapted as [i]. The reading that derived from analogy is 叉 (麻 rhyme)> 釵 [tsʰa]. The reading of 隘 is [ʌik] in HMJH and [ʌi] in YH and SG. The [ʌik] reading of 隘 is consistent with the yige 乙革 fanqie spelling in HWZY. The reading of 柴 [sɨi] is probably derived from the reading in Old Chinese (Lee 1997: 176). Lee argues that the initial of 柴 [sɨi], which is a chong 崇 initial, should had been adapted as [ts]. The [s] in [sɨi] indicates that this reading was transmitted into Korean in a very early stage, in which [ts] initial had not occurred in Korean yet. The reading of 灑 is [soai] in YH, and [sai] in SG and SH. The [soai] reading of 灑 includes an extra medial -o, which may be caused by the roundedness of the upper character 所 in the fanqie spelling suoxie 所蟹 (Synn 2006: 88). The reading of 歪 [koi] in YS version HMJH may be a copyist’s error, since it is represented as [oai] in the DD, NG, GJG and DG version of HMJH. The reading of 畫 [hoa] is borrowed from the Chinese pronunciation in Old Mandarin. In NGD and PTS, the reading of 畫 is [hoa].

Final system 121 4.4.6 Guai 夬 rhyme Table 4.14 The guai 夬 rhyme in EMSK

並 溪 影 匣 徹 崇

kaikou div. 2

hekou div. 2

Qieyun

ɣ

æi

ʷɣæi

EMSK

ai

oai

b kʰ ʔ ɦ ʈʰ dʐ

pʰai 敗

kʰoai 快

ai 餲

hoa 話i

tʰai 蠆 tsʰai 寨

i HMJH, YH, SE, SG [hoa], SU, JU, HD [hoai]

The kaikou guai 夬 rhyme is represented as [ai]. There is only two hekou 夬 rhyme characters in Table 4.14, so we have listed the representation of some hekou 夬 rhyme characters in other sources: 夬 [khoai] (IG), 噲 [khoai] (SG). These examples suggest that the hekou guai 夬 rhyme is generally represented as [oai] in MSK. The reading of 話 [hoa] is borrowed from Old Mandarin (Lee 1980: 126; Lee 1997: 177). 4.4.7 Ji 祭 rhyme Table 4.15 The ji 祭 rhyme in EMSK kaikou div. 3 Qieyun iᴇi EMSK iəi 幫 並 明 溪 疑 影 云 來 知 澄 精

p b m kʰ ŋ ʔ ɦ l ʈ ɖ ʦ

心 章 禪 書

s tɕ dʑ ɕ

iə

kaikou kaikou div. 3A div. 3B

hekou div. 3

iᴇi

ɣ

iᴇi

ʷiᴇi

iəi

əi

iəi

pʰiəi 蔽 pʰiəi 弊 miəi 袂 iəi 藝 liəi 例礪

kəi 憩揭 ui 衞i

liə 蠣 厲 勵 ii

iii

iv

tʰiəi 彘滯 tsiəi 祭穄 際 tsiəi 制製 siəi 誓 siə 噬vii筮viii逝ix siəi 世貰 siə 勢x

ui, iu, iuiəi

tʰiu 綴v

siəi 歲

tsʰiuiəi 贅vi

siəi 稅帨 iəi 蛻 (Continued)

122  Final system Table 4.15 (Continued) kaikou div. 3 Qieyun iᴇi EMSK iəi 以 j

iə

kaikou kaikou div. 3A div. 3B

hekou div. 3

iᴇi

ɣ

iᴇi

ʷiᴇi

iəi

əi

iəi

iəi 勩裔 thiəi 靾

ui, iu, iuiəi

iəi 銳睿

i HMJH, YH, SE, IG, SG, HD, SU, JU [ui] ii HMJH [liə], HD [liəi] iii HD [liəi], YH, IG, SG, SH [liə], SE [niə] iv YH [liə], HD [liəi], SE [niə] v YH [thiu], SG [tsʰiəi] vi HMJH, SG [tsʰiuəi], HD, SU, JU [tsʰiui] vii HMJH, YH [siə], IG, HD [siəi] viii HMJH, YH [siə], SE, HD [siəi] ix YH, SG [siə], SE, HD [siəi] x YH [siə], SE, IG [siəi]

The ji 祭 rhyme (kaikou div. 3) is generally represented as [iəi]. The dropping of the off-glide [i] could be a result of sound change in Korean, which was caused by the dissimilation between the on-glide [i] and the off-glide [i]. For those initials that have chongniu, 3A division is represented as [iəi], while 3B is represented as [əi]. The hekou ji 祭 rhyme is generally represented as [iəi]. The [w] glide in the source form is dropped due to the phonotactic constraints of Korean. The reading of 衞 [ui] may derived from the analogy of wei 微 rhyme characters 韋違偉 [ui] (Ito 2007: 146). The reading of 綴 is represented as [tʰiu] in YH, and [tshiəi] in SG. The reading for zhui 贅 is [tsʰiuiəi] in HMJH. The [tsʰiuiəi] reading also occurs in the hekou zhi 支 rhyme characters 嘴箠捶 [tsʰiuiəi] and the hekou zhi 脂 rhyme character 悴 [tsʰiuiəi] in HMJH. The [iui] in [tsʰiuiəi] is probably a reflection of the glide [ɥ] in Old Mandarin. 4.4.8 Fei 廢 rhyme Table 4.16 The fei 廢 rhyme in EMSK kaikou div. 3

hekou div. 3

Qieyun

iai

ʷiai

EMSK

iəi

非 敷 奉 疑 影 曉

f fʰ v ŋ ʔ h

iəi 乂i

i HMJH, YH [iəi], CJM, HD, SU, JU [ai] ii HMJH, YH, SG [ai], HD, SU, JU [iəi]

ai

uəi

iəi pʰiəi 廢 pʰiəi 肺 pʰiəi 吠

ai 刈ii huəi 喙

iəi 濊穢

Final system 123 The readings of kaikou fei 廢 rhyme characters 乂刈 in EMSK are listed as follows: 乂 HMJH, YH [iəi], CJM, HD, SU, JU [ai]; 刈 HMJH, YH, SG [ai], HD, SU, JU [iəi]. The [ai] reading may be derived from the analogy of the tai 泰 rhyme character 艾 [ai] (Lee 1997: 179). The hekou fei 廢 rhyme is represented as [uəi] after [h] initial, and [iəi] elsewhere. In addition to [huəi] 喙 listed in Table 4.16, the [uəi] reading is also seen in 顪 [huəi] (SU, JU). 4.4.9 Qi 齊 rhyme Table 4.17 The qi 齊 rhyme in EMSK kaikou div. 4

hekou div. 4

Qieyun

ei

ʷei

EMSK

iəi

疑 影 曉 匣

p pʰiəi 嬖閉 pʰ b pʰiəi 陛梐 m k kiəi 薊髻笄 雞計繫繼 kʰ kiəi 稽啓契 鸂溪 ŋ iəi 猊霓鯢詣 ʔ iəi 曀翳 h hiəi 醯 ɦ kiəi 系蹊

端

t

透

tʰ

定

d

泥 來

n l

幫 滂 並 明 見 溪

精 清

tiəi 邸觝 䏄帝堤 羝 tʰiəi 蔕 蝃嚏諦 tʰiəi 涕體  tiəi 梯 tiəi 弟第悌 娣鵜蹄啼 稊綈荑 踢𧀰提 踶 tʰiəi 髢 棣締逮

liəi 鱧醴禮棙 儷荔犁盠 鸝藜 iəi隷 ʦ tsiəi 霽濟擠 韲虀齎 ʦʰ tsʰiəi 砌

i

iə

ɨi, ʌ, ʌi iəi

pi 篦蓖螕鎞箅 pʰi 批i pi 鼙ii mi 米iii迷iv㴹

kiəi 桂v kiu 閨vi 圭vii

tiə 低底詆柢抵

ni 泥xii

iu

hʌi 嵇viii hiəi 慧ix kiu 畦xi 惠x 携

liə 麗戾沴黎 tsʌ 泲xiii tsʰiə 妻xiv 凄 (Continued)

124  Final system Table 4.17 (Continued) kaikou div. 4

hekou div. 4

Qieyun

ei

ʷei

EMSK

iəi

從 心 i ii iii iv v vi vii viii ix x xi xii xiii xiv xv xvi xvii xviii xix

ʣ tsiəi 薺臍 蠐齊 s siəi 洗細

i

iə

ɨi, ʌ, ʌi iəi

iu

siə 壻xv西xvi棲xvii犀xviii sɨi 嘶xix

HMJH [pʰi], HD [piəi] (popular reading [pi]) HMJH [pi], HD [piəi] (popular reading [pi]) HMJH, YH [mi], HD [miəi] (popular reading [mi]) HMJH [mi], HD [miəi] (popular reading [mi]) HMJH, YH [kiəi] HMJH, YH, SU, JU [kiu], HD [kiui] YH, SE, IG, SG [kiu], HD [kiui] CJM, OP [hʌi] HMJH, YH [hiəi] HMJH, NE, SE, IG, SG, SH [hiəi] HMJH, YH [kiu], HD [hiui] (popular reading [kiu]), SU, JU [hiu] HMJH [ni], HD [niəi] (popular reading [ni]) HMJH, SG [tsʌ] HMJH, YH, SU, JU [tsʰiə], HD [tsʰiəi] HMJH, YH [siə], HD [siəi] HMJH [siə], HD [siəi] HMJH, YH, SU, JU [siə], HD [siəi] HMJH, SG [siə], HD [siəi] HMJH, SU, JU [sɨi], HD [siəi] (popular reading [sɨi])

The fourth division kaikou qi 齊 rhyme is generally represented as [iəi] in EMSK. The [i] reading occurs in some kaikou qi rhyme characters when the initial is 幫 [p], 滂 [pʰ], 並 [b], 明 [m] or 泥 [n]. The qi 齊 rhyme has merged with the 支 脂之微 rhymes in the Northern Song Dynasty (Zhou 1966: 611). In ZYYY, the qi rhyme is placed under the qiwei 齊微 rhyme group, the phonetic value of which is [i]. One may view the [i] reading as a new borrowing from Chinese; however, the [i] reading is seen only in limited initials, and thus not likely to be a new layer. As Kōno (1979: 413) and Ito (2007: 142–143) note, the [i] reading in the following characters can be explained by analogy: 毘膍 (脂 rhyme) > 篦蓖螕鎞[pi], 屁秕 ( 脂 rhyme) > 批 [pʰi], 尼 (脂 rhyme) > 泥 [ni], 卑禆 (支 rhyme) > 鼙 [pi]. The [mi] reading of 米, which is a frequently used character in colloquial speech, cannot be explained by analogy. It may be a faithful reflection of the source language, in which the reading of 米 has already changed to [mi] (Ito 2007: 143). The [iə] reading in 妻壻西棲犀 may be derived from the dissimilation between the on-glide -i and the off-glide -i. The reading of 嵇 [hʌi] may be a result of hypercorrection. The reading of 泲 [tsʌ] may derived from the analogy of the zhi 脂 rhyme character 姊 [tsʌ]. The reading of 嘶 [sɨi] may derived from the analogy of the zhi 支 rhyme character廝 [sɨi]. The fourth division hekou qi rhyme is represented as [iəi]. The [iu] reading is derived from analogy: 跬 (hekou zhi 支 rhyme) > 圭畦閨 [kiu]. In HMJH, there is an annotation for 畦 [kiu], which says that “the standard reading for 畦 should be the same as 奚 [kiəi].”

Final system 125 4.4.10 Summary Below is a summary of the mainstream reading of the xie 蟹 rhyme group in EMSK. Table 4.18  Mainstream representation of the xie 蟹 rhyme group in EMSK Kaikou

div. 1 咍

泰

皆

Qieyun

əi

ɑi

ɣ

ɛi

div. 2 佳 夬 ɛ

ɣ

ɣ

æi

div. 3 祭 3/3A

3B

廢

div. 4 齊

iai

ei

iəi

iəi

iᴇi

iᴇi

ɣ

EMSK

ʌi

ai

ʌi

ai

ai

Hekou

灰

泰

皆

佳

夬

祭

廢

齊

Qieyun

uoi

w

ɛi

ʷɣɛ

ʷɣæi

ʷiᴇi

ʷiai

ʷei

oai

oai

iəi

iəi

iəi

EMSK

Labial

Other

ʌi

oi

ɑi

oi

wɣ

oi

iəi

əi

Pak (1971: 134) argues that there is a contrast between hai 咍 and tai 泰 rhymes in EMSK, suggesting it is a reflection of a Chinese variant in which hai 咍 and tai 泰 had not merge into one type. Kōno (1979: 455) claims that although hai 咍 is generally represented as [ʌi] and tai 泰 as [ai], there is no strict contrast between hai and tai 泰 rhymes in EMSK. Kōno argues that given that the kaikou tai 泰 rhyme is mainly represented as [ai], we would expect the hekou tai 泰 rhyme to be [oai]. However, it is represented as [oi] in EMSK. This shows that there is no parallel relationship between hai 咍 and tai 泰 in both kaikou and hekou. As Kōno points out, the representation of the hekou hai 咍 and tai 泰 are generally represented as the same type [oi]. However, if the source form of hai 咍 and tai 泰 is exactly the same, then we need to explain why there is a contrast between the kaikou hai 咍 and the kaikou tai 泰 rhymes in EMSK. Moreover, the [oi] representation can be explained by assuming the hekou form of tai 泰 rhyme in the source language was [wɑi]. Since the [ɑ] in [wɑi] is a back [ɑ], and due to the influence of the -w glide, [wɑi] was likely to be perceived as [woi]. Additionally, in the parallel first division double rhymes tan 覃 rhyme and tan 談 rhyme, which end with a [m] coda, we can find that some tan 覃 rhyme characters are represented with [ʌm], such as 蚕 [tsʌm]. This [ʌm] reading, however, is not seen in tan 談 rhyme characters. Therefore, it is reasonable to claim that there is a contrast between hai 咍 rhyme and tai 泰 rhyme in EMSK.

4.5  Zhi 止 rhyme group 4.5.1 Zhi 支 rhyme 4.5.1.1  Kaikou zhi 支 rhyme The kaikou 支 rhyme is generally represented as [i] in EMSK. The [ʌ] reading occurs when the initial is 精 [ʦ], 清 [ʦʰ], 從 [ʣ], 心 [s] or 日 [ȵ]. There are [ɨi] readings in a few kaikou 支 rhyme characters. Kōno (1979) claims that the [ɨi] reading

Table 4.19 The kaikou zhi 支 rhyme in EMSK kaikou div. 3

kaikou div. 3A

kaikou div. 3B

Qieyun

ie

ie

ɣ

EMSK

i

i

ɨi

幫

p

滂 並

pʰ b

明

m

見 溪 群 疑 影 曉 來

k kʰ g ŋ ʔ h l

知 徹 澄

ʈ ʈʰ ɖ

精

ʦ

清

ʦʰ

從

ʣ

心

s

崇 生 章

dʐ ʂ tɕ

昌 禪

tɕʰ dʑ

書

ɕ

船 日 以

ʑ ȵ j

ʌ

ɨi, iəi

ie

pi 卑屄髀 臂裨 pi 脾婢避

li 詈璃籬 鱺離罹 ti 智蜘知 tsʰi 黐 ti 池篪 tʰi 馳豸

tsʰi 漬

tsi 紙咫肢 枝搘 梔止支 tsʰi 巵 tsʰi 眵侈 si 鍉匙氏 是豉 i弛 si 豕翅絁 施始

mi 獼麛瀰 弭彌 ki 枳 kɨi 徛覉寄 ki 企蚑 kɨi 欹綺踦 ki 歧 kɨi 錡騎奇 ɨi 蟻義儀宜議 ɨi 椅倚 hɨi 戲曦

tsʌ 紫髭 tsʰʌ 此 tsʌ 莿雌刺 tsʌ 骴疵 sɨi 廝i sʌ 璽賜撕𪆁 斯徙 sʌ 俟 sʌ 使 sɨi 簁釃

zi 爾邇 zʌ 兒 i 匜椸扅 移迤

i HMJH [sɨi], SH [sʌ] ii HMJH [tiəi], HD, SU, JU [si]

tiəi 䑛ii

i pʰi 彼陂 pi 碑羆 pʰi 鈹披 pʰi 被鞁髲 疲皮 mi 糜𪎭靡 縻 ki 庋 ki 技妓芰

Final system 127 is a layer reflecting the Chinese reading in the Tang Dynasty, and the [ʌ] reading is a reflection of the [ɿ/ɿ] in the Song Dynasty. This was doubted by Pak (1971), who suspected that the [ʌ] reading is a reflection of Old Chinese. This study agrees that the [ɨi] is a reflection of the source form in MC. Due to the influence of alveolar/ retroflex fricative/affricate initial, a glide [i̯ ] is added before the vowel. This caused the source form to be represented as [ɨi] in SK. The origin of the [ʌ] reading is more complicated, and a detailed discussion on this issue will be provided in Chapter 6. The reading of 䑛 [tiəi] is caused by the analogy of the reading of 氐底低 [tiəi]. The 3A division zhi 支 rhyme is represented as [i]. The 3B division zhi 支 rhyme is generally represented as [ɨi] when the initial is velar or guttural, as [i] when the initial is labial. When the initial is [k], both [ki] and [kɨi] can be found. The [ki] reading should be a reflection of the sound change [kɨi] > [ki] in Middle Korean. 4.5.1.2  Hekou zhi 支 rhyme Table 4.20 The hekou zhi 支 rhyme in EMSK hekou div. 3 Qieyun

w

EMSK

iu

見 溪 群 疑 影 曉 云 來 澄 精 心 邪 初 章

k kʰ g ŋ ʔ h ɦ l ɖ ʦ s z tʂʰ tɕ

昌 禪 日

tɕʰ dʑ ȵ

hekou div. 3A

ie

w

i, ui, iui, iuiəi, iə, iəi

ie

siu 髓 siu 隨

tsʰiu 吹 siu 睡垂

li 羸iii

wɣ

ie

iu, ɨi

uəi

kiu 𨾚規 kiu 窺跬

kuəi 攱詭

ɨi 恚ii hiu 觽 liu 累 tʰiu 錘

hekou div. 3B

kuəi 跪 huəi 毀燬

ui, iu

hiu 虧i ui 危僞 ui 餵萎委 hui 麾 ui 為

 tsʰiuiəi 嘴iv tsʰiuiəi 揣v tsʰiuiəi 箠vi 捶vii tsʰiui 炊viii siə 瑞ix iəi 蘂x

i YH, CJM, SE, IG, SG [hiu], HD [kiui] (popular reading [hiu]), OP [kiu] (popular reading [hiu]) ii YH [ɨi], HD [hui] (popular reading [əi]), OP [huəi] (popular reading [əi]) iii HMJH [li], HD [lui] (popular reading [li]) iv HMJH [tsʰiuiəi], HD, SU, JU [tsʰiui] v YH [tsʰiuiəi], HD [tsʰiui] (popular reading [tsʰiu]) vi HMJH [tsʰiuiəi], SU, JU [tsʰiu], HD [tsʰiui] (popular reading [tsʰiu]) vii HMJH [tsʰiuiəi], SU, JU [tsʰiu], HD [tsʰiui] (popular reading [tsʰiu]) viii  HMJH (YS edition) [tsʰiui], HMJH (DD edition) [tsʰiu], YH [tsʰiu], SU, JU [tsʰiui], HD [tsʰiui] (popular reading [tsʰiu]) ix YJ [siəi], SH, YH, SE [siə], HD [siui] (popular reading [siə]), OP [siu] (popular reading [siə]) x HMJH [iəi], HD [zui] (popular reading [iəi])

128  Final system The hekou zhi 支 rhyme is generally represented as [iu]. The [iui] of the [iuiəi] reading is probably a mimic of the glide [ɥ] in Old Mandarin. The reading of 羸 [li] is probably derived from analogy, although it is not clear which character is the source of analogy. The 3A division hekou zhi 支 rhyme is represented as [iu]. The reading of 恚 is [ɨi] in YH, [hui] (popular reading [əi]) in HD, and [huəi] (popular reading [əi]) in OP. The [hui] and [huəi] reading is consistent with the fanqie hugui 胡桂 in HWZY, but the origin of the [ɨi] reading is not clear. The 3B division hekou zhi 支 rhyme is generally represented as [uəi] or [ui]. The reading [uəi] and [ui] in EMSK can be viewed as results of different adaptation of the same input. A similar situation can be found in the transcription of Mandarin in Korean. The Mandarin -ui [wei] after a consonant initial is transcribed as [we.i] in Choe-Kim (1985), and [u.i] in the Korean government transcription (KGT) and Eom (2002): for example, Mandarin [kwei] > [k*we.i] (Choe-Kim 1985), [ku.i] (KGT), [k*u.i] (Eom 2002). 4.5.2 Zhi 脂 rhyme 4.5.2.1 Kaikou zhi 脂 rhyme Table 4.21 The kaikou zhi 脂 rhyme in EMSK kaikou div. 3 Qieyun i EMSK i 幫 p

ʌ, ʌi

kaikou div. 3A

kaikou div. 3B

i

ɨi

i pi 匕妣 庇畀比 pʰi 秕 pʰi 屁 pi 䊧 pi 鼻琵毘 枇i pʰi 鈚膍

滂 pʰ 並 b 明 m 見 k 溪 群 疑 影 云 定 來

kʰ g ŋ ʔ ɦ d l

ki 棄 ki 鬐耆 i 蛜伊 ti 地 li 履痢蜊 梨 ni 利

ʌi

ɨi

i

əi, ʌi, uəi

pi 鄙轡 秘閟 悲 pɨi 圮ii mʌi 寐iii kɨi 肌冀 飢覬 kɨi 器 kɨi 跽 ɨi 劓 ɨi 懿

pi 備 mi 眉 麋湄 媚美

mʌi 魅iv kuəi 机v

əi 饐

Final system 129 kaikou div. 3 Qieyun i EMSK i 知 ʈ 徹 ʈʰ 澄 ɖ 娘 ɳ 精 ʦ

tʰi 致 tʰi 絺 tʰi 雉稚緻 ti 墀遲 ni 尼

清 ʦʰ 從 ʣ 心 s 邪 z 生 ʂ 章 tɕ 昌 tɕʰ 禪 dʑ 書 ɕ 船 ʑ 日 ȵ 以 j

tsi 旨指砥 脂𧹛至 祗 tsʰi 痓鴟 si 視 ki 嗜 si 矢屎屍 鳲蓍 si 示 zi 樲二貳 i 夷姨

ʌ, ʌi

kaikou div. 3A

kaikou div. 3B

i

ɨi

i

ʌi

ɨi

i

əi, ʌi, uəi

tsʌ 姊資姿 恣諮 tsʌi 齍 tsʰʌ 次 tsʌ 自瓷餈 sʌ 死四肆 私 sʌ 師獅螄

i It is [phi] in the GJ edition of CJM. ii HMJH, YH, SE [pɨi], HD [pi] (popular reading [pɨi]), SU, JU [pi] iii HMJH, YH, CJM, SG, SH [mʌi], SU, JU [mi], HD [mi] (popular reading [mʌi]) iv YH [mʌi], HD, OP [mi] (popular reading [mʌi]) v Xiamen 廈門 dialect (colloquial pronunciation [kui], literary pronunciation [ki])

The kaikou zhi 脂 rhyme is generally represented as [i]. The zhi 脂 rhyme is represented as [ʌ] where its initial is an alveolar fricative or affricate, except for the case of 齍 [tsʌi], which can be said to be derived by the analogy of the jie 皆 rhyme character 齋 [tsʌi]. The 3A division kaikou zhi 脂 rhyme is represented as [i]. The reading of 寐 [mʌi] is derived from the analogy of the hui 灰 rhyme characters 妹昧 [mʌi]. As for the 3B division, it is generally represented as [ɨi]. All the [pɨi] reading in Middle SK changed to [pi] in Modern SK. The reading of 鄙轡 [pi] is a reflection of this change. The reading of 魅 [mʌi] is also derived from the analogy of the hui 灰 rhyme characters 妹 昧 [mʌi]. According to Pak (1971: 191) the reading of 机 [kuəi] is derived from the analogy of the reading of 䡄 [kuəi], which is a hekou zhi 旨 rhyme character.

130  Final system 4.5.2.2 Hekou 脂 rhyme Table 4.22 The hekou zhi 脂 rhyme in EMSK hekou div. 3

hekou div. 3A

hekou div. 3B

Qieyun

ʷi

ʷi

ʷ ɣi

EMSK

iu

iu, iəi

uəi

u, ui, iu, oi ku 晷ii kui 龜iii koi 愧iv kiu 逵viii

oi, ui, iui, iuiəi, io

見

k

kiəi 季

kuəi 軌i

群

g ɦ l

kuəi 蕢v櫃vi饋vii

云 來

kiu 葵揆 kiəi 悸

知 澄 精

ʈ ɖ ʦ

清 從

ʦh ʣ

心

s

邪

z

生

ʂ

siu 荽邃 iu 綏 siu 繸燧篲 穗遂 siu 帥

章

tɕ

tsʰiu 萑

昌 禪 書 日 以

tɕʰ dʑ ɕ ȵ j

tsʰiu 推 siu 誰脽 siu 水 ziu 緌 iu 壝唯維惟遺

iu 帷鮪 liu 壘淚虆 niu 類 tʰiu 追 tʰiu 墜

ui 位ix

tʰoi 椎x tsʰiui 醉xi tsʰui 翠xii tsʰiuiəi 悴xiii萃xiv

tsʰoi 榱xv soi 衰 tsʰio 隹xvi錐xvii

i YH, SG, JY [kuəi], HD [kui] (popular reading [kuəi]) ii  HMJH, YH [ku], SU, JU [kuəi], HD [kui], OP [kwəi] (standard reading [kui], popular reading [ku]) iii HMJH, YH, HD, SU, JU [kui] iv YH, SE, SG, JY [koi] v HMJH [kuəi], HD [kui] (popular reading [kuəi]) vi HMJH [kuəi], HD [kui] (popular reading [kuəi]) vii YH, SE, IG, SG [kuəi] viii HMJH [kiu], HD [kiui] (popular reading [kiu]) ix HMJH, CJM, YH, NE, SE, IG, SG, JY [ui] x HMJH [tʰoi], HD [tsʰiui] (popular reading [tsʰiu]), SU, JU [tsʰiu], OP [tsʰiu] (popular reading [tʰoi]) xi HMJH, YH, HD, SU, JU [tsʰiui] xii HMJH [tsʰui], YH, CJM, HD, SU, JU [tsʰiui] xiii HMJH, YH [tshiuiəi], SU, JU [tsʰui], HD [tsʰui] (popular reading [tsʰiuəi]) xiv YH, SE, IG [tshiuiəi], SG [tshiuəi], HD [tsʰui] (popular reading [tshiuəi]) xv HMJH, HD, SU, JU [tsʰoi] xvi HMJH [tsʰio], SU, JU [tshiu] xvii HMJH [tsʰio], YH, SU, JU [tshiu], HD [tshiui] (popular reading [tshiu])

Final system 131 The 3/3A division hekou zhi 脂 rhyme is generally represented as [iu]. The onglide [i] in the reading of 醉 [tsʰiui] is a result of coarticulation. The initial [tsʰ] causes the nuclear vowel being more fronted and rounded, and, thus, an [i] ending is inserted. The readings that derived from analogy include 推 (hui 灰 rhyme) > 椎 [tʰoi], 焦蕉鷦礁樵憔 (xiao 宵 rhyme) > 隹錐 [tsʰio], 缞 (hui 灰 rhyme) > 榱 [tsʰoi] (Pak 1971: 192; Lee 1997: 222; Ito 2007: 156–157). The reading of 悴萃 [tsʰiuiəi] is probably a reflection of the reading in Old Mandarin. The [w] glide in the source form is not seen in the reading of 季悸 [kiəi]. This is consistent with the reading in Mandarin [tɕi], which is also unrounded. The 3B division hekou 脂 rhyme is generally represented as [uəi] or [ui], which are results of different adaptation strategies. The readings that derived from analogy include 咎 (you 尤 rhyme) > 晷 [ku], 魁傀塊 (hui 灰 rhyme) > 愧 [koi], 圭閨 (hekou qi 齊 rhyme) > 逵 [kiu] (Lee 1997: 222; Ito 2007: 158). 4.5.3 Zhi 之 rhyme Table 4.23  The zhi 之 rhyme in EMSK kaikou div. 3 Qieyun

ɨ

EMSK

i

見

k

溪 群

kʰ g

疑 影 曉 來

ŋ ʔ h l

知 徹 澄

ʈ ʈʰ ɖ

娘 精

ɳ ʦ

從 心

ʣ s

邪

z

莊 初 崇

tʂ tʂʰ dʐ

ki 其

li 鯉娌理里裏 吏狸嫠 ni 李 tʰi 置 tʰi 癡恥 tsi 祉 tʰi 痔峙治值 ti 持 ni 你

ɨi

ʌi

kɨi 箕基紀 記 hɨi 姬 kɨi 起欺 kɨi 棋旗麒 錤忌期 ɨi 疑擬 ɨi 意薏醫 hɨi 蟢禧喜

tʰʌi 笞i

si 葈

tsʰi 廁vi si 柿vii

ʌ

tsʰɨi 輜iii鲻iv

tsʌ 子滋孳耔 镃鼒兹 tsʌ 字牸慈鶿 sʌ 笥司鸶蕬絲 思伺 sʌ 姒耜祀飼 寺祠詞辭 嗣似 sʌ 𨧫

tsʌi 梓ii

tsʌi 滓v

sʌ 士事俟仕 (Continued)

132  Final system Table 4.23 (Continued) kaikou div. 3 Qieyun

ɨ

EMSK

i

生 章

ʂ tɕ

昌

tɕʰ

禪

dʑ

書 日

ɕ ȵ

以

j

i ii iii iv v vi vii

ɨi

ʌ

ʌi

sʌ 史使

tsi 址沚趾痣 志芝止之 tsʰi 齒蚩幟 熾嗤 si 市䦙時蒔塒 侍恃 si 詩始試 zi 耳栮餌珥㮌 輀i而 i 苡頤飴貽苢 異怡以

HMJH [tʰʌi], HD, SU, JU [tshi], OP [tshi] (popular reading [tʰʌi]) HMJH [tsʌi], HD, SU, JU [tsʌ] HMJH [tsʰɨi], HD [tsʰi] (popular reading [tsʰɨi]) HMJH [tsʰɨi], HD [tsʰi] HMJH, YH [tsʌi], HD, SU, JU [tsʌ], OP [tsʌ] (popular reading [tsʌi]) HMJH (YS edition) [tsʰi], HMJH (DD edition) [tsʰʌ], YH [tsʰʌ], HD, SU, JU [tsʰi] HMJH, YH [si]

The zhi 之 rhyme is generally represented as [ɨi], [i] and [ʌ] in EMSK according to the place of articulation. It is represented as [ɨi] when the initial is velar or guttural, [ʌ] for alveolar fricative/affricate and retroflex fricative/affricate and [i] for other initials. The readings that result from analogy include 台 (tou 透 initial hai 咍 rhyme) > 笞 [tʰʌi], 宰 (jing 精 initial hai 咍 rhyme) > 梓滓[tsʌi]. 4.5.4 Wei 微 rhyme Table 4.24 The wei 微 rhyme in EMSK kaikou div. 3

hekou div. 3

Qieyun

ɨi

w

EMSK

ɨi

i

ʌi, ɨl, oi

pi 榧篚沸扉 非飛匪 pi 妃 pi 翡剕腓肥 mi 尾味薇微

pɨl 疿i pʌi 緋ii

非

f

敷 奉 微 見

fʰ v ɱ k

溪 群 疑 影

kʰ g ŋ ʔ

kɨi 蟣機譏 鞿幾饑既 kɨi 氣豈 kɨi 璣畿祈 ɨi 毅 ɨi 扆衣依

ɨi ui

kui 鬼貴歸

oi 巍 oi 畏

ui 魏 ui 威蝛慰

uəi

Final system 133 kaikou div. 3

hekou div. 3

Qieyun

ɨi

w

EMSK

ɨi

i

曉

h

hɨi 豨希稀

云

ɦ

i ii iii iv

ɨi ʌi, ɨl, oi

ui

uəi

hui 暉楎輝 諱揮

huəi 卉iii huəi 虺iv

ui 葦蝟緯 胃韋幃闈 煒謂偉違 圍渭 hui 彙

Modern SK [pi] HMJH [pʌi], HD, SU, JU [pi] HMJH, SE, SG [huəi], HD [hoi] (popular reading [huəi]) HMJH, SE, SG, SU, JU [huəi], HD [hoi]

The wei 微 rhyme is generally represented as [ɨi] or [i], according to the place of articulation. It is represented as [i] for labial initials and [ɨi] when the initial is velar or guttural. The reading of 疿 [pɨl] is derived from the analogy of its phonetic component 弗 [pɨl]. The reading of 緋 [pʌi] is derived from the analogy of the jie 皆 rhyme characters 俳排 [pʌi]. The hekou wei 微 rhyme is generally represented as [ui] or [uəi], which are results of different adaptation strategies. The reading of 巍 [oi] is consistent with the fanqie spelling wuhui 吾回 in Jiyun and HWZY. The reading of 畏 [oi] is consistent with the fanqie wuhui 吾回 in Jiyun and the fanqie wukui 烏魁 in HWZY. 4.5.5 Summary Table 4.25 is a summary of the mainstream reading of the zhi 止 rhyme group in EMSK. The [ʌ] reading of the zhi 止 rhyme group, which will be discussed in Chapter 6, is attributed to a reflection of Old Mandarin in this study. It is noteworthy that the input [wi] in the hekou zhi 支 and hekou zhi 脂 rhyme is generally represented as [iu]. This is probably due to the phonotactic constraints in Korean. [ui] is usually only seen in syllable types [ui], [kui] and [hui] in EMSK. Since most initials of the hekou zhi 支 and hekou zhi 脂 rhyme are not velar or guttural, [wi] was adapted as [iu] in SK.

4.6  Xiao 效 rhyme group 4.6.1 Hao 豪 rhyme The hao 豪 rhyme is generally represented as [o]. The readings that derived from analogy include 條 (蕭 rhyme) > 縧 [thio], 尞燎僚 (宵 rhyme) > 潦轑 [lio], 漻 廖寥 (蕭 rhyme) > 醪 [lio]. The annotation for 轑 [lio] in HMJH states that “the standard reading is the same as 老 [lo].” The reading of 嫂艘 [su] is derived from the analogy of 叟瞍 [su]. The reading of 牢 is [loi] in HMJH, [lo] in YH, SG, HD,

Table 4.25  Mainstream representation of the zhi 止 rhyme group in EMSK 脂

支 div. 3/3A Qieyun

EMSK

ie

div. 3B ie

i

ɣ

Alveolar, Retroflex Velar other Fricative Guttural and Affricate i

div. div. 3B 3/3A

ʌ

ɨi

Qieyun

w

EMSK

ie

iu

div. 3

div. 3

ɨ

ɨi

i

ʌ

ɨi

The same as left i

ɨi

other

Velar Guttural

i

ɨi

ʌ

脂

支 div. 3/3A

微

ɣ

The same as left

i

之

div. 3B wɣ

uəi

ie

div. div. 3B 3/3A w

ui

微

i

iu

wɣ

div. 3 w

i ui

uəi ui

uəi

Table 4.26 The hao 豪 rhyme in EMSK kaikou div. 1 Qieyun

ɑu

EMSK

o

幫 並 明 見 溪 疑

p b m k kʰ ŋ

影 曉 匣 端 透

ʔ h ɦ t tʰ

定 泥 來 精 清

d n l ʦ ʦʰ

從

ʣ

po 堡褓寶報保‍pʰo 褒 pʰo 暴菢袍抱 mo 帽毛瑁旄耄冒媢 ko 槀‍誥‍告‍羔‍膏‍橰‍櫜‍篙‍餻‍皐高藁 ko 槁‍拷‍考‍‍𥬯‍犒‍尻 o‍ 鏊螯‍熬‍厫‍獒傲 ‍ko ‍ 翱‍‍ o 襖‍媼‍爊奧 ho 好‍‍薅‍蒿 mo 耗 ho 昊‍號‍豪‍壕‍嘷‍浩毫灝皓 to 島‍擣‍祷‍刀‍舠倒到叨 tʰo‍ 套討 ‍to ‍ 饕韜 to 道‍稻‍盜‍桃‍陶‍萄‍淘‍濤逃導悼蹈 no 瑙‍腦 lo 老‍䳓‍䇭‍澇‍撈‍𤛮勞 tso 藻‍早蚤澡棗‍‍𥨫糟躁遭 tsʰo ‍草‍騲‍ tso‍ 糙操 tso 皂‍造‍槽‍曹‍艚‍螬

io

oi

thio 縧 nio 臑i lio 潦ii轑醪iii

loi 牢iv

ɨi

u

Final system 135 kaikou div. 1

心 i ii iii iv v vi

Qieyun

ɑu

EMSK

o

s

‍‍tso‍ 噪臊‍燥 so‍ 搔繅騷掃

io

oi

u su ‍嫂v艘vi

HMJH [nio], HD, SU, JU [no] HMJH [lio], SG, SU, JU [lo] HMJH, HD [lio], SU, JU [lo], OP [lo] (standard reading [lio]) HMJH [loi], YH, SG, HD, SU, JU [lo], OP [lo] (popular reading[loi]) HMJH [su], HD, SU, JU [so], OP [so] (popular reading [siu]) HMJH [su], HD, JU [so]

SU, JU, [lo] (popular reading [loi]) in OP. The [loi] reading includes an extra offglide [i], which may arise from the influence of the following grammatical word, such as the nominative case marker -i and the predicative case marker -i(ra) (Kōno 1979: 447). 4.6.2 Yao 肴 rhyme Table 4.27 The yao 肴 rhyme in EMSK kaikou div. 2

幫 滂 並 明 見 溪 影 曉 匣 知 澄 娘 心 莊 初 崇 生 1 ii iii iv v vi

Qieyun

ɣ

EMSK

o

io

p pʰ b m k kʰ ʔ h ɦ ʈ ɖ ɳ s tʂ tʂʰ dʐ ʂ

pʰo 飽苞‍包 pʰo 砲‍疱‍泡‍脬‍‍抛 pʰo 皰‍庖‍匏‍咆‍炮‍跑‍麅‍齙‍鉋 mo 貌‍蝥‍茅‍

pʰio 豹‍i

æu

ho‍ 髇‍‍ii to‍ 棹‍‍iii tso 爪iv‍笊‍ so‍ 巣‍‍vi so 鞘

oa

kio ‍鉸‍狡‍酵‍校‍教‍窖‍覺‍咬‍交‍膠‍郊攪‍ kio 磽‍骹‍巧 io‍ 坳 lio ‍‍拗 hio ‍孝‍哮 hio ‍肴‍‍效 tio‍ 嘲‍‍ nio 鬧‍‍鐃‍‍ tsʰio 炒‍鈔‍ tsʰio 梢‍艄稍 弰‍蛸‍筲‍‍

HMJH, YH, HD, SU, JU [pʰio] HMJH [ho], HD [hio] HMJH, YH [to], HD, OP [tso] (popular reading [to]), SU, JU [tso] HMJH, YH, SG [tso] HMJH (YS edition) [koa], HMJH (DD edition) [tsoa], HD, SU, JU [tso] HMJH, YH, SE, IG, SG, HD [so], OP [tsʰio] (standard reading [so]), SU, JU [tsʰio]

koa‍ 抓‍‍v

136  Final system The yao 肴 rhyme is represented as [o] or [io]. The distribution of [o] and [io] is generally according to the place of articulation. The [o] form mainly occurs for labial initials, while [io] elsewhere (the [o] reading is a reflection of the sound change [io] > [o] for fricative/affricate initials in Middle Korean). The reading of 抓 [koa] is derived from the analogy of the ma 麻 rhyme character 瓜 [koa]. 4.6.3 Xiao 宵 rhyme Table 4.28 The xiao 宵 rhyme in EMSK kaikou div. 3

kaikou div. 3A

kaikou div. 3B

Qieyun

iᴇu

iᴇu

ɣ

EMSK

io

io

io

pʰio 標 pʰio‍ 瓢‍薸‍鳔縹剽飄漂 mio 緲眇妙渺 tsʰio 杪

pʰio‍ 表‍臕‍鑣‍攩‍

幫 並

p b

明

m

見 溪 群 影 云 來 知 徹 澄 精 清 從 心

k kh g ʔ ɦ l ʈ ʈʰ ɖ ʦ ʦʰ ʣ s

章

tɕ

昌 禪 書 日 以

tɕʰ dʑ ɕ ȵ j

kio‍ 蹺‍ ‍io 腰邀 lio ‍燎‍療 tio‍ 朝‍ tʰio 超 tio 朝‍潮‍‍肇兆趙 sio召 tsʰio‍‍ 椒‍焦‍礁‍蕉‍鷦醮 tsʰio ‍‍锹‍ tsʰio ‍‍誚‍樵‍憔 sio ‍笑宵霄‍焇鞘小消銷逍 ‍tsʰio ‍綃‍ sio ‍沼昭 tsio 照‍詔‍𨹸 tʰio 招 tsʰio‍ 麨‍tʰio ‍弨‍‍ sio 紹邵劭 sio 燒‍少 io 蕘饒擾繞遶 zio 橈 io ‍曜‍鷂‍窰‍瑤‍軺‍謠‍舀遙搖耀

iᴇu

mio 廟‍苗‍描‍貓 kio‍ 嬌矯驕 kio 鞽 kio 轎‍橋‍蕎 io 夭‍妖 hio ‍鴞‍‍

The xiao 宵 rhyme is a third division rhyme. It is represented as [io] in EMSK for both the 3/3A division and the 3B division. 4.6.4 Xiao 蕭 rhyme The xiao 蕭 rhyme is a fourth division rhyme. It is represented as [io] in EMSK, implying that a [j] glide has occurred in the source form.

Final system

137

Table 4.29 The xiao 蕭 rhyme in EMSK kaikou div. 4

見 溪 疑 影 曉 端 透 定 泥 來 心

Qieyun

eu

EMSK

io

k kʰ ŋ ʔ h t tʰ d n l s

kio 徼叫皎 hio 梟 kiu 竅 io 僥 mio 杳 lio 窈 hio 曉 tio 鳥釣雕錭彫弔凋 tʰio 貂 tio 眺糶 tio 銚跳蜩鰷條調 tʰio 齠 nio 尿 lio 蓼鐐僚鷯聊遼瞭了寥 nio料 sio 嘯簫蟏蕭

4.6.5 Summary The mainstream representation of the xiao 效 rhyme group in EMSK is summarized in Table 4.30. Table 4.30 Mainstream representation of the xiao 效 rhyme group in EMSK 豪

肴

宵

蕭

kaikou div. 1 kaikou div. 2 kaikou div. 3/3A ɣ au iᴇu Qieyun ɑu EMSK o Labial Other io o io

kaikou div. 3B ɣ iᴇu io

kaikou div. 4 eu io

4.7 Liu 流 rhyme group 4.7.1 Hou 侯 rhyme Table 4.31 The hou 侯 rhyme in EMSK kaikou div. 1 Qieyun

u

EMSK

u

o

滂 並 明 見

pʰ b m k

pu 剖 pu 部抔裒 mu 貿楙茂瞀袤 ku 垢笱狗構購覯彀雊 溝冓篝鉤韝耈苟

mo 某i牡ii拇iii母iv畝v姆vi

溪

kʰ

ku 口釦簆鷇寇 彄摳

ko 叩

iu

(Continued )

138  Final system Table 4.31 (Continued) kaikou div. 1 Qieyun

u

EMSK

u

疑 影 曉 匣

ŋ ʔ h ɦ

端 透 定

t tʰ d

泥 來

n l

u 藕腢偶 ku 嘔漚甌鷗謳蓲 hu 吼 hu 後后鍭候堠侯猴糇 喉帿逅厚 tu 斗抖蚪𥆖陡 tʰu 闘 tʰu 偷透 tu 痘荳脰餖頭竇豆 tʰu 骰㢏䬦投

精 清 心

ʦ ʦʰ s

o

iu

to 篼

tiu 鍮vii niu 㝅viii

lu 耬瘻漏鏤簍螻髏摟 蔞樓 陋 tsu 奏湊走 tsu 腠湊 su 擞瞍叟薮嗽漱

i HMJH, YH [mo], HD, SU, JU, OP [mu] ii HMJH, YH, SE, SG [mo], HD, SU, JU [mu], OP [mu] (popular reading[mo]) iii HMJH [mo], IG, HD, SU, JU [mu], OP [mu] (popular reading [mo]) iv  HMJH, HG, CJM, YH, NE, SE, IG, SG, JY, SH [mo], HD, SU, JU [mu], OP [mu] (popular reading [mo]) v  HMJH, YH [mo], CJM, SE [mio], SG [mo], [mio], HD, SU, JU [mu], OP [mu] (popular reading [mio]) vi HMJH [mo], SH [mu] vii HMJH, HD [tiu], SU, JU [thu], OP [tiu] (popular reading [iu]) viii HMJH [niu], HD, SU, JU [nu]

The hou 侯 rhyme is generally represented as [u]. Huang (1931: 105) points out that in qinyin 秦音 (Chang’an dialect) in Huilin 慧琳’s (737–820) time, the hou 侯 characters mu 拇, mu 母 and mu 牡 had already merged into the mu 暮 rhyme. The [o] reading had probably occurred in those characters in the source language. The reading of 鍮 [tiu] and 㝅 [niu] is probably derived from analogy although it is not clear which character is the source of analogy. The reading of 叩 [ko] and 篼 [to] is probably a reflection of Old Chinese, in which the phonetic value of the hou 侯 rhyme is [o] according to the reconstruction of Wang Li and Lu Zhiwei (Lee 1997: 246). 4.7.2 You 尤 rhyme Table 4.32 The you 尤 rhyme in EMSK kaikou div. 3 Qieyun ɨu EMSK u 明 m 非 f

iu

o

iui, iuk

mo 矛 眸 鍪 麰 謀 i

pu 缶殕否缹富𦉓

ii

iii

iv

kaikou div. 3 Qieyun ɨu EMSK u 敷 奉 見 溪 群

fʰ v k kʰ g

疑 影 曉 云 來

ŋ ʔ h ɦ l

知 ʈ 徹 ʈʰ 澄 ɖ 娘 ɳ 精 ʦ 清 ʦʰ 從 ʣ 心 s 邪 莊 初 崇 生 章

z tʂ tʂʰ dʐ ʂ tɕ

昌 tɕʰ 禪 dʑ 書 ɕ 日 ȵ 以 j

i ii iii iv v vi vii viii ix x

iu

pu 副覆 pu 婦負阜蝜蜉芣𣑿浮 ku 九韭廄鳩鬮久救究 ku 糗丘蚯 ku 舅臼柩裘仇毬梂求 舊咎 u牛 u 優耰麀憂 hu 朽嗅 hiu 鵂休 u 右友祐疣尤 iu 囿有宥又 liu 柳畱瘤溜餾 琉旒榴鶹流留 tiu 肘晝輈 tʰiu 抽 tiu 宙胄紬籌疇 儔稠 niu 紐鈕扭糅 liu 杻狃 tsiu 酒 tsʰiu 秋鰍楸湫 鞦鶖䐐鞧 tsʰiu 鷲 siu 繡銹饈脩 秀羞 siu 袖岫泅囚 tsʰu 皺縐 tsʰiu 甃vi緅vii tsʰu 篘 su 愁 su 瘦蒐溲螋搜 tsʰiu 箒 tsiu 洲州舟周賙 siu 受壽售酬 讎綬 siu 手首狩獸 收守 ziu 葇 iu 柔揉 iu 牖莠誘柚游蚰 油蝣斿猶由攸 猷輶

o

iui, iuk

tsʰiui 僦v tsʰiui 就

tsʰiui 驟 tsʰiuk 呪viii tsʰiui 醜ix 臭x

HMJH, SE, SG [mo], HD, OP [mu] (popular reading [mo]), SU, JU [mu] HMJH [mo], HD, OP [mu] (popular reading [mo]), SU, JU [mu] HMJH [mo], HD [mu] (popular reading [mo]), SU, JU [mu] HMJH, YH [mo], HD, OP [mu] (popular reading [mo]), SU, JU [mu] HMJH [tsʰiui], HD, SU, JU [tshu], OP [tshu] (popular reading [tsʰiui]) HMJH, IG, HD [tsʰiu], SU, JU [tshu] HMJH [tshiu], SU, JU [tshu] HMJH [tsʰiuk], HD, SU, JU [tsu] HMJH, HG, YH, IG, SG [tsʰiui], HD, SU, JU [tshu], OP [tshu] (popular reading [tsʰiui]) HMJH, YH, SG, SG, JY [tsʰiui], HD, SU, JU [tshu], OP [tshu] (popular reading [tsʰiui])

140  Final system The you 尤 rhyme is generally represented as [u] or [iu] according to the place of articulation. It is represented as [u] when the initial is labial, velar, guttural or a retroflex fricative/affricate, and [iu] elsewhere. The reading of 矛眸鍪麰謀 [mo] is probably a faithful representation of the source language, in which the readings of these characters have merged into the mu 模 rhyme. According to Huang (1931: 108), in qinyin 秦音 (Chang’an dialect), the you 尤 rhyme characters 矛 謀 had already merged into mu 模 rhyme in Huilin 慧琳’s (737–820) time. The reading of 呪 [tsʰiuk] is derived from the analogy of the wu 屋 rhyme character 祝 [tsʰiuk] (Pak 1971: 182; Synn 2006: 95). The reading of 僦醜臭就驟 [tsʰiui] is a result of coarticulation. The initial [tsʰ] causes the nuclear vowel being more fronted and rounded, and, thus, an [i] ending is inserted. 4.7.3 You 幽 rhyme The you 幽 rhyme is a third division rhyme. Here are lists of the readings of some you 幽 rhyme characters in EMSK: 繆 [liu], 糾 [kiu], 幼幽 [iu]. This shows that the representative form of the you 幽 rhyme is [iu] in EMSK. 4.7.4 Summary Based on the above discussion, the mainstream representation of the liu 流 rhyme group can be summarized in Table 4.33. Table 4.33  Mainstream representation of the liu 流 rhyme group in EMSK

Qieyun EMSK

侯

尤

幽

kaikou div. 1 u

kaikou div. 3 ɨu labial, guttural, retroflex fricative/ affricate u

kaikou div. 3 iu

u

other iu

iu

4.8  Xian 咸 rhyme group 4.8.1 Tan 覃 rhyme Table 4.34 The tan 覃 rhyme in EMSK

見 溪 影 匣

覃 (kaikou div. 1)

合 (kaikou div. 1)

Qieyun

əm

əp

EMSK

am

k kʰ ʔ ɦ

kam 紺感敢 kam 坎墈堪 ham 龕 am 暗庵䳺諳 ham 撼頷含函涵 kam 憾

ʌm

ap hap 鴿閤蛤 ap 罨 hap 䤴合

Final system 141

端 透 定 泥 來 精 清 從 心

覃 (kaikou div. 1)

合 (kaikou div. 1)

Qieyun

əm

əp

EMSK

am

t tʰ d n l ʦ ʦʰ ʣ s

tam 湛 tʰam 耽 tham 探貪黮 tam 潭蟫壜藫覃 nam 男南楠 nam 嵐 lam 婪 tsam 揝 tsʰam 參慘 tsham 糝ii

ʌm

ap tap 答 tap 錔踏 nap 納衲 tsap 咂匝

tsʌm 蚕i

tsap 雜 sap 靸

i  HMJH [tsʌm]. The traditional form of the character 蚕, namely 蠶, is represented as [tsʌm] in YH, SE, HD and [tsam] in SG, SU, JU. ii HMJH (YS edition) [tsham], HMJH (DD edition), HD [sʌm], SU, JU [sam]

The tan 覃 rhyme is generally represented as [am], except in the cases of 蚕 [tsʌm] (HMJH) and 糝 [sʌm] (HMJH (DD edition), HD). The he 合 rhyme is represented as [ap] in EMSK. Pak (1971: 141) and Kōno (1979) claim that this [ʌm] reading is a reflection of Old Chinese, arguing that the tan 覃 rhyme is reconstructed as [əm] in Karlgren’s reconstruction of Old Chinese. However, in the hai 咍 rhyme, another first division rhyme parallel to the tan 覃 rhyme, we can see both [ʌi] and [ai] readings. This is in parallel to the [ʌm] and [am] in the tan 覃 rhyme. Given that we attribute the [ai] reading of the hai 咍 rhyme to a result of sound change from [ʌi] in Middle Korean, this study views the [am] reading of the tan 覃 rhyme as a result of the sound change [ʌm] > [am] in Middle Korean. 4.8.2 Tan 談 rhyme Table 4.35 The tan 談 rhyme in EMSK 談 (kaikou div. 1)

盍 (kaikou div. 1)

Qieyun

ɑm

ɑp

EMSK

am

見 溪 曉 匣 端 透 定 來

k kʰ h ɦ t tʰ d l

從 心

ʣ s

kam 甘柑泔 kam 瞰坩 ham 憨 kam 酣 tam 膽擔 tam 毯菼 tam 噉淡談痰澹 lam 攬爁覽纜濫 籃襤 nam 藍 tsam 暫 tsʰam 慙 sam 三

an

ap hap 榼 hap 闔

tan 坍

thap 塔搭榻塌 nap 臘鑞 lap 蝋 sap 𢻨

142  Final system The tan 談 rhyme and the he 盍 rhyme are represented as [am] and [ap] respectively. The reading of 坍 [tan] is derived from the analogy of the reading of its phonetic component 丹 [tan]. 4.8.3 Xian 咸 rhyme Table 4.36 The xian 咸 rhyme in EMSK 咸 (kaikou div. 2)

見 溪 影 匣 知 澄 莊 初 崇 生 i ii iii iv

洽 (kaikou div. 2)

Qieyun

ɣ

EMSK

am

iəm

k kʰ ʔ ɦ ʈ ɖ tʂ tʂʰ dʐ ʂ

kam 減 ham 緘

kiəm 鹻i

ɛm

ɛp

ɣ

am 黯 ham 餡陷鹹咸 tsʰam 站 tam 賺 tsam 蘸 tsʰam 斬

ap

iəp

ɨp

hiəp 裌ii夾 kiəp 掐iii tsap 劄

hiəp 狹

hɨp 洽iv

sap 鍤插 tsap 牐煠 sap 歃箑

tsʰam 讒饞 sam 釤杉

HMJH, HD, SU, JU [kiəm] HMJH [hiəp], HD [kiəp] (popular reading [hiəp]), SU, JU [kiəp] HMJH [kiəp], HD [kap] HMJH, YH, SE, SG [hɨp], HD, OP [hiəp] (popular reading [hɨp])

The xian 咸 rhyme is generally represented as [am]. The reading of 鹻 [kiəm] is derived from the analogy of the reading of its phonetic component 兼 [kiəm], which is a fourth division tian 添 rhyme character. The qia 洽 rhyme is generally represented as [ap]. The reading of 裌 [hiəp] is probably derived from the analogy of 俠頰挾 [hiəp], which are fourth division tie 帖 rhyme characters. The final of 洽 [hɨp] is the same as the reading of ji 緝 rhyme, which is generally represented as [ɨp]. The reading of 洽 [hɨp] is consistent with the fanqie huji 胡急 in Yunbu 韻 補 [Supplementary guide to rhymes], a rhyme book compiled by Wu Yu 吴棫 (ca. 1100–1154) in the Song Dynasty. 4.8.4 Xian 銜 rhyme Table 4.37 The xian 銜 rhyme in EMSK 銜 (kaikou div. 2)

見 溪 疑

狎 (kaikou div. 2)

Qieyun

ɣ

ɣ

EMSK

am

ap

k kʰ ŋ

kam 鑑監 kam 嵌 am 巌

kap 甲胛

æm

æp

Final system 143 銜 (kaikou div. 2)

影 匣 生

狎 (kaikou div. 2)

Qieyun

ɣ

ɣ

EMSK

am

ap

ʔ ɦ s

ham 艦檻銜 sam 衫

æm

æp

ap 鴨壓 kap 匣 ap狎

The xian 銜 rhyme and the xia 狎 rhyme are represented as [am] and [ap], respectively, in EMSK. 4.8.5 Yan 鹽 rhyme Table 4.38 The yan 鹽 rhyme in EMSK 鹽 kaikou div. 3/3A

葉 kaikou div. 3B

kaikou div. kaikou div. 3/3A 3B

iᴇm

iᴇp

ɣ

iəp

iəp

Qieyun iᴇm

ɣ

EMSK iəm, ʌm

əm

幫 見 群

p k g

kəm 瞼檢 kəm 儉黔

疑 影

ŋ ʔ

iəm 黶檿𢻨厭

曉 云

h ɦ

həm 險 iəm 炎

來

l

知 徹

ʈ ʈʰ

娘 精

ɳ ʦ

liəm 殮臉iii 廉匳簾帘鎌臁 𧸘歛 tiəm 沾 tʰiəm 諂 tiəm 覘

清 從

ʦʰ ʣ

心

s

tsʰiəm 尖 siəm 殲 tsʰiəm 籤憸 tsiəm 漸 tsʌm 潛iv siəm 纖

həm 驗 əm 閹醃晻奄掩

iəm, am, ɨm

iᴇp

pʰiəm 貶 kam 芡i kiəm 鉗 kɨm 鈐ii am俺

niəp 饁燁 liəp 鬛獵 niəp 躐

niəp 鑷 tsʰiəp 睫 tsiəp 椄接 tsʰiəp 妾 tsʰiəp 捷

(Continued )

144  Final system Table 4.38 (Continued) 鹽 kaikou div. 3/3A

葉 kaikou div. 3B

kaikou div. kaikou div. 3/3A 3B

iᴇm

iᴇp

ɣ

iəp

iəp

Qieyun iᴇm

ɣ

EMSK iəm, ʌm

əm

章

tɕ

昌 禪 書

tɕʰ dʑ ɕ

日 以

ȵ j

iəm, am, ɨm

tsʰiəm 瞻 tsiəm 占 tsʰiəm 韂 siəm 蟾贍 siəm 苫閃 tsiəm 痁 ziəm 染髯枏 iəm 燄扊艷閻鹽 焰豔 ziəm tsʰiəm 簷

iᴇp

siəp 涉 siəp 韘攝 iəp 葉

i HMJH [kam], HD, SU, JU [kəm], OP [kəm] (popular reading [kam]) ii YH [kɨm], HD [kiəm] iii Modern SK [kəm] iv  HMJH, YH, CJM, SE, IG, SG, JY, HD [tsʌm], SU, JU [tsʰiəm], OP [tsʰiəm] (standard reading [tsʌm])

The yan 鹽 rhyme is generally represented as [iəm]. The 3/3A division yan 鹽 rhyme is generally represented as [iəm] except in the case of 潛 [tsʌm]. The reading of 潛 [tsʌm] may be derived from the reading of tan 覃 rhyme character 簪 [tsʌm] (Ito 2007: 173). The 3B division yan 鹽 rhyme is generally represented as [əm]. The reading of 芡 [kam] is derived from the analogy of the tan 覃 rhyme character 坎 [kam] (Pak 1971: 147). The reading of 鈐 is [kɨm] in YH, and [kiəm] in HD. The [kɨm] reading of 鈐 is probably derived from the analogy of its phonetic component 今 [kɨm]. An 俺 is represented as [am] in HMJH. The fanqie for an 俺 in Guangyun 廣韻 is yuyan 於驗; therefore, the expected SK reading for an 俺 is [əm]. The reading of 俺 [am] is probably derived from the analogy of the reading of tan 覃 rhyme character 庵 [am] (Pak 1971: 147; Ito 2007: 174). The ye 葉 rhyme is represented as [iəp] for both the 3/3A division and the 3B division. 4.8.6 Yan 嚴 rhyme Table 4.39 The yan 嚴 rhyme in EMSK

見 溪 疑

嚴 (kaikou div. 3)

業 (kaikou div. 3)

Qieyun

iam

iap

EMSK

əm

k kʰ ŋ

kəm 劍 əm 釅嚴儼

ɨm

əp

hɨm 欠i

kəp 劫 kəp 怯 əp 業

Final system

影 曉

145

嚴 (kaikou div. 3)

業 (kaikou div. 3)

Qieyun

iam

iap

EMSK

əm

ɨm

əp

ʔ h

əm 淹 həm 蘝

hɨm 杴ii

həp 脇脅

i HMJH, YH [hɨm], HD [kiəm] (popular reading [hɨm]), OP [kəm] (popular reading [hɨm]), SU, JU [kəm] ii HMJH [hɨm], HD, SU, JU [həm]

The yan 嚴 rhyme and the ye 業 rhyme are mainly represented as [əm] and [əp], respectively, in EMSK. The reading of 欠杴 [hɨm] may derived from the analogy of the qin 侵 rhyme character 欽 [hɨm] (Kōno 1979: 464). 4.8.7 Tian 添 rhyme Table 4.40 The tian 添 rhyme in EMSK

見 溪 匣 端 透 定 泥 精 心

添 (kaikou div. 4)

帖 (kaikou div. 4)

Qieyun

em

ep

EMSK

iəm

k kʰ ɦ t tʰ d n ʦ s

kiəm 蒹縑兼 kiəm 歉謙慊 hiəm 嫌 tiəm 店點 tʰiəm 舔添忝 tiəm 簟㼭 tʰiəm 甜 niəm 恬 niəm 鲇念 tsiəm 拈

ʌm

iəp hiəp 莢頰筴鋏蛱 hiəp 篋愜 hiəp 協挾俠

tsʰʌm 僭ii

tʰiəp 帖貼 tiəp 蝶楪 tʰəp 牒i tʰiəp 疊 niəp 捻 hiəp 浹 siəp 爕

i HMJH [tʰəp], YH, CJM [tʰiəp], SH [tshiəp] ii HMJH, YH, SE, SG, HD [tshʌm], OP [tsiəm] (standard reading [tshʌm])

The tian 添 rhyme and the tie 帖 rhyme are represented as [iəm] and [iəp], respectively, in EMSK. The reading of 牒 is represented as [tʰəp] in HMJH, [tʰiəp] in YH and CJM and [tshiəp] in SH. The reading of 僭 [tsʰʌm] may be derived from the reading of 譖 [tsʰʌm]. 4.8.8 Fan 凡 rhyme Table 4.41 The fan 凡 rhyme in EMSK 凡 (hekou div. 3)

非

Qieyun

w

EMSK

əm f

iɐm

乏 (hekou div. 3) w

iɐp

əp, ip pəp 法 (Continued )

146  Final system Table 4.41 (Continued) 凡 (hekou div. 3)

乏 (hekou div. 3)

Qieyun

w

EMSK

əm

əp, ip

pəm 泛 pəm 帆凡犯範

pʰip 乏i

fʰ v

敷 奉

iɐm

w

iɐp

i HD [phiəp] (popular reading [phip]), YH, SH [phip], OP [pəp] (popular reading [phip])

The fan 凡 rhyme and the fa 乏 rhyme are represented as [əm] and [əp], respectively, in EMSK. The reading of 乏 [pʰip] is irregular, and the origin of this reading is not clear. 4.8.9 Summary Below is a summary of the mainstream representation of the xian 咸 rhyme group in EMSK. Table 4.42  Mainstream representation of the xian 咸 rhyme group in EMSK 覃

談

咸

銜

鹽

kaikou kaikou kaikou kaikou 3/3A div. 1 div. 1 div. 2 div. 2 ɣ ɣ am am iem Qieyun əm ɑm EMSK am am am am iəm

3B iem əm ɣ

嚴

凡

添

kaikou div. 3 iem əm

hekou div. 3 w iɐm əm

kaikou div. 4 iem iəm

4.9  Shen 深 rhyme group 4.9.1 Qin 侵 rhyme Table 4.43 The qin 侵 rhyme in EMSK kaikou div. 3

kaikou div. 3B

Qieyun im EMSK 幫 滂 見

p pʰ k

溪

kʰ

群

g

疑

ŋ

im

ɣ

ɨm

ʌm

im

ɨm

kɨm 錦今 襟金禁 kɨm 衾 hɨm 欽 kɨm 妗禽 琴檎噙 ɨm 吟

am

um pʰum 稟i pʰum 品ii

Final system 147 kaikou div. 3

kaikou div. 3B

Qieyun im EMSK 影

ʔ

曉 來

h l

知

ʈ

澄 娘 精 清 從 心 邪 莊

ɖ ɳ ʦ ʦʰ ʣ s z tʂ

崇 生 章

dʐ ʂ tɕ

昌 禪

tɕʰ dʑ

書 船 日

ɕ ʑ ȵ

以

j

i ii iii iv

im

lim 霖林臨 sim 椹 tʰim 砧 tʰim 沈 nim 賃 tsʰim 浸 tsʰim 寢䈜侵 sim 蕈 sim 心 sim 燖鬵尋

tsʰim 枕鍼 tsim 斟 sim 瀋 tʰim 忱 sim 甚 sim 嬸深審 sim 葚 nim 稔恁 zim 𥙛絍任 妊荏餁 im 紝

im

ɣ

ɨm

ʌm

lɨm 廩iv檁 凜

ɨm

am

ɨm 飲窨 音陰蔭 hɨm 歆

am 瘖iii

um

tsʰʌm 譖 tsʌm 簪 tsʌm 涔梣 sʌm 蔘森滲 tsʌm 箴

ɨm 淫

YH, SE [phum] HMJH, YH, SE, IG, SH [phum], IG, HD, SU, JU [pɨm] HMJH [am], HD, SU, JU [ɨm] HMJH, YH, HD, SU, JU [lɨm]

The 3/3A division qin 侵 rhyme is generally represented as [im]. The reading of 廩檁凜 is [lɨm] in EMSK. The fanqie for lin 廩檁凜 is lijin 力錦 in Jiyun. Given that the reading of jin 錦 is [kɨm] in EMSK, the reading of 廩檁 [lɨm] is consistent with this fanqie in Jiyun. The 3/3A division qin 侵 rhyme is represented as [ʌm] when the initial is retroflex zhuang 莊 [tʂ], chong 崇 [dʐ] or sheng 生 [ʂ]. For the [ʌm] reading, Pak (1971: 194) argues that its origin can be traced to Old Chinese, in which the phonetic value of the qin 侵 rhyme is [ïəm]. However, in Guangyun there are two fanqie spellings for zan 簪 (zuohan 作含 and ceyin 側吟) and shen 蔘 (suhan 蘇含 and suojin 所今). The reading of 簪 [tsʌm] and 蔘 [sʌm] may be a reflection of the first fanqie spelling. The reading of 森 [sʌm] is consistent with the fanqie shuzan 疏簪, and the reading of 滲 [sʌm] is consistent with fanqie

148  Final system shuzan 疎簪 in Jiyun. The reading of 涔梣 [tsʌm] is consistent with the fanqie chuzan 鉏簪 in Jiyun. The 3B division qin 侵 rhyme is generally represented as [ɨm]. The reading of 瘖 [am] is derived from the analogy of the reading of the tan 覃 rhyme characters 暗諳 [am]. The reading of 稟品 [pʰum] should be the result of the sound change [pʰɨm] > [pʰum]. 4.9.2 Ji 緝 rhyme Table 4.44 The ji 緝 rhyme in EMSK

見 溪 群 影 曉 來 澄 清 從 邪 莊 生 章 禪 書 日

kaikou div. 3

kaikou div. 3A

kaikou div. 3B

Qieyun

ip

ip

ɣ

EMSK

ip

ɨp

ɨp

k kʰ g ʔ h l ɖ ʦʰ ʣ z tʂ ʂ tɕ dʑ ɕ ȵ

ɨp

ɨp 揖i挹ii lip 粒笠立苙 tʰip 蟄 tsip 集

tsip 執 sip 十 ip 入

ip

kɨp 給級汲急 ɨp 泣 kɨp 及 ɨp 邑浥 hɨp 吸

tsɨp 緝iii tsɨp 楫iv輯 sɨp 習v襲 tsɨp 蕺戢 sɨp 澁 sɨp 拾 sɨp 濕

i HMJH, YH, NE, SG [ɨp] ii HMJH, YH, SG [ɨp] iii HMJH, YH, SG, HD, SU, JU [tsɨp] iv HMJH, YH, SE, SG [tsɨp] v HMJH, YH, CJM, NE, SE, IG, SG, SH, HD, SU, JU [sɨp]

The third division ji 緝 rhyme is represented as [ip] or [ɨp]. The [ɨp] reading occurs after a [+strident] initial. This can be explained as a result of sound change in SK. The mix of [ɨ] and [i] after [ts]/[tsh]/[s] initial is seen in Middle SK, such as 瑟 [sɨl] > [sil], 什 [sɨp] > [sip], 叱 [tsɨl] > [tsil], 輯 [tsɨp] > [tsip], 則 [tshɨp] > [tship], 襯齔 [tshɨn] > [tshin] (Nam 1973: 125). The 3A division ji 緝 rhyme is represented as [ɨp] in 揖挹 [ɨp]. The initial ying 影, which is a glottal [ʔ], may causes the [ip] being perceived as [ɨp] by borrowers. The 3B division ji 緝 rhyme is represented as [ɨp]. 4.9.3 Summary The mainstream representation of the shen 深 rhyme group is listed in Table 4.45.

Final system 149 Table 4.45  Mainstream representation of the shen 深 rhyme group in EMSK 侵 Qieyun EMSK

緝

kaikou div. 3(A) im im

kaikou div. 3B im ɨm

ɣ

kaikou div. 3(A) ip ip ɨp

kaikou div. 3B ip ɨp

ɣ

4.10  Shan 山 rhyme group 4.10.1 Han 寒 rhyme and huan 桓 rhyme Table 4.46 The han 寒 rhyme in EMSK 寒

曷

桓

kaikou div. 1

kaikou div. 1 hekou div. 1

Qieyun ɑn

ɑt

w

EMSK an

al

an

幫 p 滂 pʰ 並 b 明 m

oan, ən, uən al

tʰal 奪

kan 看刊 an 岸 an 案鞍安按

曉 h

han 罕漢

匣 ɦ

han 旱悍銲鼾 汗翰俧寒 捍韓 tan 担 疽旦箪褝丹單 tʰan 炭灘攤擹 嘆坦 tʰan 袒彈誕憚 tal 薘達 tan 壇檀 lan 難 lan 嬾瀾欄蘭 lal 辣糲 懶爛

泥 n 來 l

ɑt

tan 段團椴

溪 kʰ 疑 ŋ 影 ʔ

定 d

ɑn

pal 鉢 pal 醱

kan 稈笴幹干 肝竿奸漧 han 旰趕

透 tʰ

hekou div. 1 w

pan 般半 pʰan 判 pən 潘i pan 冸 pan 伴盤 瘢蟠叛磻 man 漫 幔蔓鰻鏝饅 鞔謾滿曼 kal 葛 koan 管輨 hal 割 盥館鸛罐 觀官棺冠 莞涫觀 貫灌 kal 渴 koan寬款欵 al 枿 oan 玩 al 遏閼 oan 腕豌帵 uən 盌 hoan 獾渙 歡喚煥 kal 毼 hoan 換紈芄 萑桓丸 oan 完緩澣 tal 怛 tan 鍛碬端 短斷 tal 撻獺闥韃 tan 湍猯

見 k

端 t

末

oal

pal 鈸拨 mal 抹沫 䴲末 koal 括 筈

hoal 蛞闊 al 斡 hal 豁 hoal 活

tʰal 脫

nan 暖 lan 卵鸞亂 man 巒 (Continued )

150  Final system Table 4.46 (Continued) 寒

曷

桓

末

kaikou div. 1

kaikou div. 1 hekou div. 1

Qieyun ɑn

ɑt

w

EMSK an

al

an

tsʰal 拶

tsʰan 鑽 tsan 纂 tsʰan 爨鑹

精 ʦ

tsʰan 讚贊

清 ʦʰ 從 ʣ 心 s

tsʰan 粲飡 tsan 殘 san 傘糤珊散

hekou div. 1

ɑn

w

ɑt

oan, ən, uən al

oal

tsʰal 撮

san 算酸狻 suən 蒜ii 餕

i HMJH [pən], HD, SU, JU [pan] ii HMJH, HD [suən], SU, JU [san]

The han 寒 rhyme and the he 曷 rhyme are represented as [an] and [al], respectively. The huan 桓 rhyme is represented as [an] or [oan] following place of articulation. It is generally represented as [an] for labial, alveolar initials and as [oan] elsewhere. The [an] reading for labial is consistent with Middle Chinese in which labial initials do not have a rounded/unrounded contrast. The [an] reading for alveolar initials is due to the phonotactic constraint in Korean. This is supported by the evidence from the transcription of Mandarin in Korean. Eom (1996: 77) claims that the Korean government transcription includes many monosyllabic representations which are rarely used in native Korean phonology, such as [twɑn], [thwɑn] etc. In Eom (1996), these syllable types are avoided by appealing to the resyllabification strategy. The reading of 潘 is [pən] in HMJH, and represented as [pan] in HD, SU, JU. The [pən] reading is derived from the analogy of the yuan 元 rhyme characters 番藩 [pən]. The [uən] reading is probably a reflection of the Chinese reading in Old Mandarin. The notation of 蒜 [suən] and 盌 [uən] in HMJH is the same as the Chinese readings given in SSTH, NGD and PTS. The mo 末 rhyme is represented as [al] or [oal] according to the place of articulation. It is represented as [al] for labial, alveolar initials and as [oal] elsewhere. 4.10.2 Shan 山 rhyme Table 4.47 The shan 山 rhyme in EMSK

幫 並 見

山

黠

山

黠

kaikou div. 2

kaikou div. 2

hekou div. 2

hekou div. 2

Qieyun

ɣ

ɣ

wɣ

EMSK

an, ʌn

al, il

oan

p b k

pan 扮 pʰan 瓣辦 kan 簡㶕襇間艱揀

pʰal 八

ɛn

ɛt

al 鴶 kal 圿

ɛn

hoan 鰥

wɣ

ɛt

oal

Final system 151 山

黠

山

黠

kaikou div. 2

kaikou div. 2

hekou div. 2

hekou div. 2

Qieyun

ɣ

ɣ

wɣ

EMSK

an, ʌn

al, il

oan

oal

溪 疑 匣

kʰ ŋ ɦ

hil 黠

hoan 幻

hoal 猾滑

澄 莊 初 崇 生

ɖ tʂ tʂʰ dʐ ʂ

kan 慳 an 眼 kan 癇 han 閑 hʌn 限i tʰan 綻 tsan 盞剗

ɛn

ɛt

tsan 棧孱 san 產𩥮山疝

ɛn

wɣ

ɛt

tsʰal 察 tsʰal sal 殺

i YH, HD, IG [hʌn]

The kaikou shan 山 rhyme and the hekou shan 山 rhyme are represented as [an] and [oan], respectively. The reading of 限 [hʌn] in YH, HD and IG may be derived from the analogy of the hen 痕 rhyme character 恨 [hʌn]. The kaikou xia 黠 rhyme and the hekou xia 黠 rhyme is represented as [al] and [oal], respectively. The reading of 黠 is [hal] in JU and [hil] in YH. The final [il] in the [hil] reading may arise from the influence of the phonetic radical 吉 [kil] (Lee 1997: 153). 4.10.3 Shan 刪 rhyme Table 4.48 The shan 刪 rhyme in EMSK 刪

鎋

kaikou div. 2

kaikou div. 2 hekou div. 2

Qieyun ɣæn EMSK an 幫 p 滂 明 見 疑 影 曉 匣

pʰ m k ŋ ʔ h ɦ

泥 初 崇 生

n tʂʰ dʐ ʂ

æt

oan

pʰan 板 pan 螌頒 pan 攀 man 蠻慢 kan 澗諫 koan 菅 an 贗鴈顏 an 晏

wɣ

al

oan

koan 關慣 oan 頑 hal 瞎 hal 轄

lan 赧 san 鏟

tsʰal 剎

鎋 hekou div. 2

æn

ɣ

han 骭

san 訕刪

刪

wɣ

an

lan 薍 man 彎i

æt

oal, oa

koal 刮

hoan 宦豢槵環 寰鬟還患

soan 孿𣟴

i HMJH [man], HD, OP [oan] (popular reading [man]), SU, JU [oan] ii HMJH, YH, NE [tshan] iii HMJH [soa], YH [sal], HD, SU, JU [soal]

tsʰan 饌ii撰

soa 刷iii

152  Final system The kaikou shan 刪 rhyme is represented as [an]. The reading of 菅 [koan] is derived from the analogy of the reading of the huan 桓 rhyme characters 官 管 [koan]. The hekou 刪 rhyme is represented as [oan]. The reading of 薍 [lan] is derived from the analogy of the reading of 亂 [lan]. The reading of 彎 [man] is probably derived from the analogy of the reading of the ming 明 initial character 蠻 [man] (Pak 1971: 50). The kaikou xia 鎋 rhyme and the hekou xia 鎋 rhyme are represented as [al] and [oal], respectively. The hekou xia 鎋 rhyme character shua 刷 is represented as [soa] in HMJH. The native Korean annotation for this character is [soa. tsʌ], which is a loanword from Chinese shuazi 刷 子 “brush.” The [t] coda in MC is always consistently represented as [l] in SK. This [soa] reading is apparently of Old Mandarin origin, as shown by the loss of [t] coda. 4.10.4 Xian 仙 rhyme Table 4.49 The xian 仙 rhyme in EMSK kaikou kaikou kaikou div. 3 div. 3A div. 3B Qieyun iᴇn

iᴇn

ɣ

EMSK iən

iən

ən

幫

p

滂

pʰ

並

b

明

m

pʰiən 鞭 編 pʰiən 騙 篇偏 pʰiən 便 楩 miən 面 綿

見

k

溪

kʰ

群

g

疑 影 云

ŋ ʔ ɦ

來

l

知

ʈ

澄

ɖ

hekou div. 3

iᴇn

w

iən

iᴇn

iən

w

iᴇn

wɣ

iᴇn

iən

uən

kiən 絹

kuən 眷卷

oan

piən 變

piən 辯弁辨 miən 娩 冕免 勉 kiən 甄 kən kiən 謇瀽 囝 kiən 譴 kən 攐愆 遣 kən 乾 虔件 ən 彦 ən 焉

liən 輦鰱連 聯 tiən 𩥇展 tsiən 鳣 tiən 廛纏

hekou hekou div. 3B div. 3A

kuən 拳 koan 顴 權倦

liən 戀 tiən 轉 iən 椽 tiən 傳

uən 院衏 員圓

Final system 153 kaikou kaikou kaikou div. 3 div. 3A div. 3B Qieyun iᴇn

iᴇn

ɣ

EMSK iən

iən

ən

娘

ɳ

niən 碾

精

ʦ

清

ʦʰ

從

ʣ

心

s

邪

z

莊 章

tʂ tɕ

tsiən 剪 箭煎 湔翦 tsʰiən 韆遷淺 tsiən 餞錢 tsʰiən 踐賤 siən 癣 燹藓 獮線 仙秈 siən 羨 iən 涎

昌

tɕʰ

禪

dʑ

書

ɕ

船 日

ʑ ȵ

以

j

tsiən 戰 氈饘 鹯 tsʰiən 闡 siən 善鱔 蟮蟬 饍禪 tsʰiən 擅 siən 騸扇 tsiən 羶 iən 然 iən 蜒筵 演延 衍

hekou div. 3

iᴇn

w

iən

hekou hekou div. 3B div. 3A

iᴇn

iən

w

iᴇn

iən

wɣ

iᴇn

uən

oan

tsiən 鐫 tsiən 痊 悛 iən 吮 tsʰiən 泉 tsiən 全 siən 宣 選 siən 旋 璇 tsiən 跧 tsiən 甎 專剸 tsʰiən 喘 釧穿 川舛 tsʰiən 腨 篅

siən 船 ziən 蠕 㮕撋 iən 軟 iən 鳶鉛 沿緣 捐

The 3/3A division kaikou xian 仙 rhyme is represented as [iən]. The 3B division kaikou xian 仙 rhyme is generally represented as [ən] and as [iən] when the initial has a labial place of articulation. The reading of 囝 [kiən] is [kən] in SU and JU. The [kiən] reading is the same as the notation in SSTH, implying a close relationship with Old Mandarin (Nam 1973: 146). The 3/3A division hekou xian 仙 rhyme is represented as [iən]. The 3B division hekou xian 仙 rhyme is represented as [uən] generally. The reading of 顴 [koan] is derived from the analogy of the huan 桓 rhyme characters 鹳觀 [koan].

154  Final system 4.10.5 Xue 薛 rhyme Table 4.50 The xue 薛 rhyme in EMSK kaikou div. 3

kaikou div. 3A kaikou div. 3B

Qieyun iᴇt

iᴇt

ɣ

EMSK iəl

iəl

əl

幫 並 明 見 群 疑 來

p b m k g ŋ l

piəl 鼈虌

知 徹 澄 從 心 章 書 船 日 以

ʈ ʈʰ ɖ ʣ s tɕ ɕ ʑ ȵ j

iᴇt

hekou div. 3 w

iəl, iək

iᴇt

iəl

ol

piəl 別

miəl 滅搣 hiəl 孑

kəl 傑 əl 孽闑糵 piək 蘖i

liəl 蛚烈鴷冽 niəl 列裂 tʰiəl 哲 tʰiəl 徹 tʰiəl 轍澈

liəl 劣 tʰiəl 啜綴掇輟 tsiəl 絶 siəl 雪 tsiəl 棁 siəl 說

siəl 疶泄褻泄 tsiəl 靼折 siəl 設 siəl 舌 iəl 熱

tsol 拙ii

siəl 䑔 iəl 悅閱

i HMJH [piək], HD, SU, JU [əl] ii YH, SE, HD, OP [tsol]

The 3/3A division kaikou xue 薛 rhyme and the 3B division kaikou xue 薛 rhyme are represented as [iəl] and [əl], respectively. The 3/3A division hekou xue 薛 rhyme is represented as [iəl]. The reading of 蘖 [piək] is probably derived from the analogy of the reading of 檗薜 [piək] or 僻 [piək] (Ito 2007: 186). The reading of 拙 is [tsol] in YH, SE, HD and OP. Since the fanqie spelling of 拙 is zhiyue 職 悅 in Guangyun 廣韻, the expected reading of 拙 in SK is [tsiəl]. The [tsol] reading of 拙 is the same as mo 沒 rhyme characters 卒𣨛 [tsol]. 4.10.6 Yuan 元 rhyme Table 4.51 The yuan 元 rhyme in EMSK 元

月

元

kaikou div. 3 kaikou div. 3 Qieyun iɤn

iɤt

EMSK ən

əl

非 f

hekou div. 3 w

al

月 hekou div. 3

iɤn

w

ən, uən, un

an, oan

pən 藩蕃

pʰan 阪 販畈 pan 反返

iɤt

əl, uəl

al pal 髮發

Final system 155 元

月

元

kaikou div. 3 kaikou div. 3 Qieyun iɤn

iɤt

EMSK ən

əl

hekou div. 3 w

al

月 hekou div. 3

iɤn

ən, uən, un

w

an, oan

敷 fʰ 奉 v

pən 旙轓翻 pən 燔 pan 飯 蘩礬𢶃繁

微 ɱ

man 晚 萬挽

見 k

kən 犍鞬建

溪 kʰ 群 g

kən 鍵健

疑 ŋ

ən 言

影 ʔ

ən 堰鰋

曉 h

hən 獻軒憲

云 ɦ

kal 羯 al 訐i

iɤt

əl, uəl pəl 罰筏 瞂昁 伐

al

mal 韈

kuəl 蹶 蕨鱖瘚厥 kuəl 闕 kuəl 橛

kuən 勸券 kal 碣 kuən 圈 竭 uən 原黿芫蚖 oan 阮 uəl 月刖 元願源 al 謁 uən 苑鴛冤 oan 畹ii婉 əl 噦 踠怨 həl 蠍歇 huən 楦萱 暄貆喧 hun 塤iii uən 遠猿園垣 uəl 鉞越 轅榬援

oal 曰

i HMJH, YH [al], HD, OP [kal] (popular reading [al]), SU, JU [kal] ii HMJH [oan], HD, SU, JU [uən] iii YH, HD [hun], OP (standard reading [hun]), SU, JU [huən]

The kaikou yuan 元 rhyme is generally represented as [ən]. The hekou yuan 元 rhyme is generally represented as [uən], and [ən]/[an] when appearing after labial initials. The reading of 塤 is [hun] in YH, HD, OP (standard reading) and [huən] in SU, JU. In Jiyun it says that 塤 is a variant of the character 壎. The [hun] reading of 塤 is consistent with the fanqie spelling of 壎, namely xuyun 許云, in HWZY. The kaikou yue 月 rhyme is represented as [əl] or [al]. The hekou 月 rhyme is generally represented as [uəl], and [əl]/[al] for labial initials. The reading of 畹 [oan] is derived from the analogy of the reading of the huan 桓 rhyme characters 完碗 [oan]. Pak (1971: 156) claims that the readings of the following characters result from analogy: 曷 (he 曷 rhyme) > 羯 [kal], 扳眅 (shan 刪 rhyme) > 阪販畈 飯 [pʰan], 跋𧺺拔 (mo 末 rhyme) > 髮 [pal], 鞔悗 (huan 桓 rhyme) > 晚 [man], 瀎𥽘 (mo 末 rhyme) > 襪 [mal]. The analogy analysis proposed in Pak (1971) suffers from the following weakness. The reading of a frequently used character, such as 晚, is not likely derived from the reading of less frequently used characters, such as 鞔悗. Shin (2015: 182) explains the [ən]/[an] reading as a reflection of two different layers. The [an] reading is a reflection of the Chinese in the Southern and Northern Dynasties (420–589), and the [ən] reading a reflection of the Chinese in the Tang Dynasty. However, the [ən]/[an] contrast for the hekou yuan 元 rhyme is also seen in Sino-Vietnamese, such as 反返 [phan], 飯 [phan], 晚 [van], 萬 [van], 發髮 [phat], 伐罰 [phat], 番 [phien], 煩 [phien], 筏 [phiet] (Wang 1948: 47). In addition,

156  Final system in the transliteration of Sanskrit by Xuanzang 玄奘 (AD 602–664), the pronunciation of the hekou yuan 元 or the yue 月 rhyme in some labial initial characters have already been close to [a], such as 軬飯 for [vɑn], 伐 for [vɑt]/[vɑr]/[vɑl]/[vɑl], 筏 for [vɑt]/[vɑr]/[vɑl], 罰 for [vɑr]/[vɑs]/[vɑj] (Shi 1983: 42–43). Kōno (1979: 463–464) notices that in hekou yuan 元 rhyme, there is a tendency that for labial initials ping tone is represented as [ən], shang tone and qu tone as [an], and ru tone as either [əl] or [al]. Ito (2007: 186) notes that this tendency is generally also obeyed in Sino-Vietnamese and suggests that the glide of the hekou yuan 元 rhyme is clearly perceived when it is a ping tone but not easily perceived when it is a shang tone or a qu tone. This study agrees to Ito’s (2007) analysis and reconstructs the phonetic value for the kaikou yuan 元 rhyme in the source language as [iɐn]. 4.10.7 Xian 先 rhyme Table 4.52 The xian 先 rhyme in EMSK 先

屑

先

kaikou div. 4

kaikou div. 4

hekou hekou div. 4 div. 4

et

w

Qieyun en EMSK iən 幫p 滂 pʰ 並b 明m 見k 溪 kʰ 疑ŋ 影ʔ 曉h 匣ɦ

pʰiən 蝙 piən 邊藊遍徧 pʰiən 片 pʰiən 艑 piən 玭 miən 眠麵眄 kiən 繭襺趼毽見 肩鵑堅 hiən 筧 kiən 牽 iən 硯研妍 iən 嚥鷰宴煙胭 咽淵 hiən 蜆顯 hiən 峴莧賢弦舷 縣絃

in

iəl

iəi

tʰiən 天

定d

tiən 殿澱電淀甸 tin 殄v 奠畋癜靛田塡 niən 撚年 niəl 捏 liən 錬憐蓮鍊 niən 練 tsʰiən 薦 tsiəl tsiən 牮箋 癤節

et

iəl, iul

iəi 蠮 hil 翓iv

透 tʰ

iən

w

kiən 畎 kiəl 決 蠲 hiul iən 涓 譎iii kiən 犬 kiəl 缺

hiəl 挈 iəl 霓 hiəl 齧

tiən 典巔癲顛

精ʦ

ɨl, al

miəl 篾蠛 kiəl kil 桔i䓀ii 鍥𢎹結潔

端t

泥n 來l

il

en

屑

hɨl 齕麧

tʰiəl 餮鐵 til 垤vi 跌vii迭耋

nal 苶viii

hiəl 血 hiən 玄 hiəl 穴 衒懸 眩

Final system 157 先

屑

先

kaikou div. 4

kaikou div. 4

hekou hekou div. 4 div. 4

et

w

Qieyun en EMSK iən 清 ʦʰ 從ʣ 心s

in

tsʰiən 蒨千𢶊 tsiən 前 siən 跣霰先

iəl

il

ɨl, al

iəi

en

iən

屑

w

et

iəl, iul

tsiəl 竊切 tsiəl 截 siəl 糏楔屑

i HMJH [kil], HD, SU, JU [kiəl] ii  HMJH (YS edition) [kil]. In the annotation of HMJH (YS edition), Choe writes that “the standard reading should be [kiəl].” HD, SU, JU [kiəl] iii YH [hiul], HD, OP (popular reading) [hiul] iv HMJH [hil], HD, SU, JU [hiəl] v YH, SE, SG [tin], HD, OP (popular reading) [tin] vi HMJH [til], HD, OP [tiəl] (popular reading [til]) SU, JU [tiəl] vii HMJH [til], HD, SU, JU [tiəl], OP [tiəl] (popular reading [til]) viii YH [nal], OP (popular reading [nal]), SU, JU [niəl]

Both kaikou xian 先 rhyme and hekou xian 先 rhyme are represented as [iən]. The reading of 殄 [tin] is derived from the analogy of zhen 真 rhyme characters 珍診疹畛 [tin]. The kaikou xie 屑 rhyme and the hekou xie 屑 are represented as [iəl] except for the following readings which are derived from analogy: 吉 (質 rhyme) > 桔䓀 [kil], 翳 (齊 rhyme) > 蠮 [iəi], 詰 (質 rhyme) > 翓 [hil], 姪侄 (質 rhyme) > 垤耋 [til], 秩佚 (質 rhyme) > 跌迭 [til], 迄屹 (迄 rhyme) > 齕麧 [hɨl], 捺 (曷 rhyme) > 苶 [nal], 鷸繘 (術 rhyme) > 譎 [hiul]. 4.10.8 Summary The mainstream representation of the shan 山 rhyme group is summarized in Table 4.53.

4.11  Zhen 臻 rhyme group 4.11.1 Hen 痕 rhyme The hen 痕 rhyme is represented as [ɨn] or [ʌn]. The hun 魂 rhyme, which will be discussed later, is the hekou corresponding rhyme of the hen 痕 rhyme and is generally represented as [on] or [un] in EMSK. This is parallel to the [ʌn]/[ɨn] reading of the hen 痕 rhyme. Kōno (1979: 470–471) argues that the vowel [ʌ] (Hangul symbol “ㆍ”) in Middle Korean was [ə] in Old Korean, and, therefore, the [ə] in the source form was adapted with Old Korean [ə]. Later, as the vowel “ㆍ” in Korean underwent a shift from [ə] to [ʌ], Korean started to represent the [ən] in Chinese with [ɨn]. Kōno’s (1979) analysis is reasonable, but we cannot exclude the possibility that the [ɨn]/[ʌn] readings result from different adaptation strategies. Due to the influence of the nasal coda [n], the nuclear vowel in the hen 痕 rhyme was perceived to be closer to [ɨ], rather than [ə], thus it was adapted as [ɨ] in most cases. This is also seen in NGD and PTS, in which the reading of 根跟 is [kɨn].

158  Final system Table 4.53  Mainstream representation of the shan 山 rhyme group in EMSK 寒

山

刪

Kaikou

div. 1

Qieyun

ɑn

ɣ

EMSK

an

an

Kaikou

曷 div. 1

黠 鎋 div. 2 div. 2

ɑt al 桓 div. 1

ɣ ɛt æt al al 山 刪 div. 2 div. 2

Qieyun EMSK Hekou Qieyun EMSK other

Hekou

ɛn

w

w

other al

æn

div. 3(A) iᴇn

an

iən

ɣ

ɣ

ɑn velar/ guttural an oan 末 div. 1

Qieyun EMSK

div. 2 div. 2

ɑt

ɛn oan

wɣ

æn oan

wɣ

div. 3(A) iᴇt iəl

ɛt

oal

wɣ

元

先

div. 3B

div. 3

div. 4

iᴇn

iɤn

en

other labial ən iən 薛 div. 3B

ən

iən

月 div. 3

屑 div. 4

ɣ

iᴇt əl

仙

oal

al

元 div. 3

iᴇn uən

iɤn uən an

wɣ

w

əl

div. 3

薛 div. 3

æt

iɤt

ɣ

div. 3(A) w iᴇn iən

黠 鎋 div. 2 div. 2 wɣ

velar/ guttural

仙

iᴇt

iəl

et iəl 先 div. 4 en iən

w

w

月 div.3

屑 div.4

w

uəl

iɤt

w

al

et

iəl

oal

Table 4.54 The hen 痕 rhyme in EMSK kaikou div. 1

見 溪 影 匣 透

Qieyun

ən

EMSK

ɨn

k kh ʔ ɦ th

kɨn 根跟 kɨn 齦 ɨn 恩 hɨn 痕狠

ʌn kʌn 墾懇 hʌn 恨 tʰʌn 吞i

i  HMJH, YH, HD, SU, JU [tʰʌn]

4.11.2 Zhen 真 rhyme The zhen 真 rhyme is generally represented as [in]. 鬢賓 [piŋ] and 牝嬪 [piŋ] are represented as [pin] in YH, HD, SU and JU. The reading of 蜃 [siun] probably results from the analogy of the zhun 諄 rhyme character 脣 [siun]. 3B division is generally represented as [ɨn] and [in] for labial initials. The reading of 巾 in MSK is listed as

Table 4.55 The zhen 真 rhyme in EMSK kaikou div. 3

kaikou div. 3A

kaikou div. 3B

Qieyun

in

in

ɣ

EMSK

in

iun

in

in

iŋ

幫

p

pin 殯擯

piŋ 鬢i賓ii

並

b m

piŋ 牝iii嬪iv蘋v

明 見

k

pin 獱㰋嚬頻 min 泯民 kin 緊

溪 群 疑 影

kʰ g ŋ ʔ

曉 來

h l

知 徹 澄 精 清 從 心

ʈ ʈʰ ɖ ʦ ʦʰ ʣ s

邪 莊 章

z tʂ tɕ

禪

dʑ

書

ɕ

船 日

ʑ ȵ

以

j

i ii iii iv v vi vii

in 印姻 茵因 lin 悋䗷麟 鱗鄰吝 tin 珍鎮 tin 趁 tin 塵陳 tsin 進津晉 tsʰin 親 tsin 螓秦盡 sin 信訊顖 辛薪新 迅 sin 燼 tsin 臻 tin 疹畛診 tsin 賑真 缜振震 sin 腎臣晨宸 辰鷐慎 sin 哂身紳 娠伸申 sin 神 zin 訒刃 仁人 in 忍仞認 in 蚓引寅

ɨn

ən, in

pin 貧 min 旻敏 憫愍 kən 巾vi kɨn 僅 ɨn 銀誾 hɨn 釁

siun 蜃 vii

iun 胤

HMJH [piŋ], YH, HD, SU, JU [pin] HMJH, CJM [piŋ], YH, NE, SE, IG, SG, SH, HD, SU, JU [pin] HMJH [piŋ], SE, IG, SG, HD, SU, JU [pin] HMJH [piŋ], SE, SG [pin] HMJH [piŋ], YH, SG [pin] CJM [kin], HMJH, SG [kən], HD, OP [kɨn] (popular reading [kən]), SU, JU [kɨn] HMJH [siun], HD, SU, JU [sin]

160  Final system follows: CJM [kin], HMJH, SG [kən], HD, OP [kɨn] (popular reading [kən]), SU, JU [kɨn]. The [kən] reading of 巾 is viewed as a reflection of Old Mandarin in Lee (1997: 235). However, the notation for 巾 [kən] in SSTH is [kin], implying that the reading of 巾 [kən] is not borrowed from Old Mandarin. The reading of 巾 [kən] is probably a reflection of Old Chinese (Pak 1971: 200; Kōno 1979: 472; Shin 2015: 192). The character 巾 is a wen 文 group character in OC, which is reconstructed as [ən] in Karlgren, Wang Li and Li Fang-kuei’s reconstruction. 4.11.3 Zhi 質 rhyme Table 4.56 The zhi 質 rhyme in EMSK kaikou div. 3

kaikou div. 3A

kaikou div. 3B

Qieyun

it

it

ɣ

EMSK

il

il

ɨl

幫 滂 並 明 見 溪 影 來 知 澄

p pʰ b m k kʰ ʔ l ʈ ɖ

娘 清

ɳ ʦʰ

從 心 章

ʣ s tɕ

昌 書 日 以

tɕʰ ɕ ȵ j

til 銍窒 til 姪 帙秩 nil 昵 tsʰil 七 漆㯃 tsil 疾嫉 sil 蟋悉 til 蛭桎 tsil 礩櫍質 sil 室實失 zil 日 il 溢逸泆鎰

ɨl, iul

liul 栗i篥

pʰil 觱必畢 pʰil 匹 pʰil 柲 mil 蜜謐 kil 吉 hil 詰 il 一壹

it il pʰil 筆 mil 密

ɨl 鳦

sɨl 膝ii tsɨl 叱

i HMJH [liul], YH, SE, SH [niul] ii HMJH [sɨl], HG, YH [sil]

The 3/3A division zhi 質 rhyme is generally represented as [il]. The reading of 栗篥 [liul] results from the analogy of the shu 術 rhyme characters 律葎 [liul]. The 3B division zhi 質 rhyme is represented as [il] when the initial is labial and as [ɨl] elsewhere. 4.11.4 Zhen 臻 rhyme The zhen 臻 rhyme is represented as [in], and the zhi 櫛 rhyme is represented as [ɨl] in EMSK. Pan and Zhang (2013: 6) states that “zhen 臻 and zhi 櫛 rhyme

Final system 161 Table 4.57 The zhen 臻 rhyme in EMSK

莊 初 生

臻

櫛

kaikou div. 3

kaikou div. 3

Qieyun

ɪn

ɪt

EMSK

in

ɨl

tʂ tʂʰ ʂ

tsin 榛i tsʰin 齔ii櫬

tsɨl 櫛 sɨl 瑟蝨

i HMJH, YH, SG [tsin] ii HMJH [tsʰin], HD, SU, JU [tsʰɨn]

were in complementary distribution with zhen 真 and zhi 質 rhyme. The former only co-occur with the initial group zhuang 莊 and the latter elsewhere. When rhyme zhen 真 occurs with zhuang 莊 initial, which are retroflexes, the place of its articulation will move back.” The difference between the zhi 質 rhyme and the zhi 櫛 rhyme in MC is reflected in SK, in which the zhi 質 rhyme is represented as [il] while the zhi 櫛 rhyme is represented as [ɨl]. 4.11.5 Yin 殷 rhyme Table 4.58 The yin 殷 rhyme in EMSK

見 溪 群 疑 影 曉

殷

迄

kaikou div. 3

kaikou div. 3

Qieyun

ɨn

ɨt

EMSK

ɨn

ɨl

əl

k kh g ŋ ʔ h

kɨn 謹槿靳斤筋

hɨl 訖

kəl 吃i kəl 乞ii肐虼

kɨn 近芹勤 ɨn 垽齗 ɨn 癮殷隱 hɨn 脪昕欣

hɨl 屹

i HMJH [kəl], HD, SU, JU [kɨl] OP [kɨl] (popular reading [hɨl]) ii HMJH, YH, NE [kəl], HD, OP [kɨl] (popular reading [kəl]) SU, JU [kɨl]

The yin 殷 rhyme is represented as [ɨn], while the qi 迄 rhyme is represented as [ɨl] or [əl] in EMSK. The [ɨl] reading of the qi 迄 rhyme is also seen in other sources: 訖 [hɨl] (HD, SE), 迄 [hɨl] (SG). In a word, there are two types of readings for the qi 迄 rhyme, namely [ɨl] and [əl]. Lee (1997: 239) claims that the [əl] reading may be derived from the reading in Old Mandarin. However, the reading of 吃 and 乞 is [tʃi] and [khi] respectively in NGD and PTS (Hu 1963: 188). Following viewing the reading of 巾 [kən] as a reflection of OC, here we also consider the [əl] reading as a reflection of OC. The characters 吃乞肐虼 have [ə] as the main vowel in the OC reconstruction of Karlgren, Wang Li and Li Fang-kuei.

162  Final system 4.11.6 Hun 魂 rhyme Table 4.59 The hun 魂 rhyme in EMSK 魂

沒

hekou div. 1

hekou div. 1

Qieyun

uon

uot

EMSK

un

on

幫 滂 並

p pʰ b

pun 錛奔 pun 噴 pun 盆

pon 畚本

明

m

mun 門捫

見

k

kun 裩

溪

kʰ

疑 影 曉

ŋ ʔ h

匣

ɦ

端

t

透 定

tʰ d

泥 來 精 清 從 心

n l ʦ ʦʰ ʣ s

i ii iii iv v vi vii

tun 臀ii飩iii屯iv鈍 tsun 樽vi tsun 蹲vii

kon 棍滾昆 蜫袞鵾 kon 閫坤壼 困悃 on 搵瘟溫穩 hon 昏婚閽 惛 hon 混渾圂 魂餛 ton 頓墩驐 敦惇 ton 褪暾 ton 沌囤豚 魨遁 non 嫩 lon 論 tson 尊 tsʰon 寸村 tson 鐏存 son 孫猻蓀 飧損巽遜

an, in

man 暪 min 悶

ol

mol 沒 kol 鶻骨 榾

ul, ʌl

pʌl 鵓脖 𩓐荸

kul 窟i

ol 兀 hol 笏忽 惚 kol 𣝗 tol 柮 tol 堗突 nul 訥v tsol 卒𣨛

HMJH, YH [kul] HMJH, YH, IG [tun] HMJH [tun], HD, SU, JU [ton] HMJH, IG [tun] HMJH, NE [nul] HMJH, YH, IG [tsun] HMJH, YH, SG [tsun]

The hun 魂 rhyme is generally represented as [on] or [un]. The reading of 暪 [man] is derived from the analogy of 瞞 [man]. The reading of 悶 [min] is derived from zhen 真 rhyme characters 閩閔 [min]. The mo 沒 rhyme is generally represented as [ol]. There is no kaikou and hekou contrast for labial initials in Middle Chinese. The [ʌl] reading after labial initials should be regarded as a faithful representation of the source language. The readings of 噴 [pun], 窟 [kul], 訥 [nul] are [phun], [khu], [nu] in SSTH, indicating that they were not borrowed from Old

Final system 163 Mandarin. The [un] reading for the hun 魂 rhyme and the [ul] reading for the mo 沒 rhyme cannot be explained by sound change in Korean. If there was a sound change that [o] was raised to [u] in Korean, we should had seen [uŋ] reading for the dong 東 rhyme and the dong 冬 rhyme, which are represented as [oŋ] in Middle SK. The [un] reading for the hun 魂 rhyme may be explained by different adaptation strategies. The input form [ṷon] is generally represented as [on]; however, in some cases, especially after labial initials, it tends to be perceived as [un]. 4.11.7 Zhun 諄 rhyme Table 4.60 The zhun 諄 rhyme in EMSK 諄

術

hekou div. 3

hekou div. 3

Qieyun

w

EMSK

iun

見 溪 群 云 來 知 徹 澄 精 清 從 心

k kʰ g ɦ l ʈ ʈʰ ɖ ʦ ʦʰ ʣ s

kiun 麕鈞均 kiun 囷 kiun 菌 kiun 筠 liun 輪淪綸掄倫

邪 生 章 昌

z ʂ tɕ tɕʰ

禪 書 船

dʑ ɕ ʑ

tsiun 準稕 tsiun 蠢 tsʰiun 春 siun 踳 siun 醇蓴鶉純 siun 蕣瞬v siun 楯盾唇順

日 以

ȵ j

ziun 閏 iun 潤 iun 尹允

i ii iii iv v

in

tsʰiun 椿 tsiun 俊儁 tsiun 皴逡竣

w

un

it

iul

ol

kiul 橘 kun 窘 un 隕i tun 窀ii迍 tsun 遵

tsiun 隼峻浚 siun 筍詢

liul 葎律 tsul 茁 tʰiul 黜 tʰiul 𦬸朮

hiul 恤 siul 銊 tʰiul 訹

siun 旬巡循徇馴 tun 肫iv

tsol 崒

sol 蟀iii率 tsʰiul 出

siul 術述 tʰiul 秫 hiul 鷸繘

YH, SE, IG, SG [un] HMJH, HD [tun], OP [tsiun] (standard reading [tun]) HMJH, YH, SG, HD [sol], OP [siul] (standard reading [sol]), SU, JU [siul] HMJH [tun], JY [tsiun] HMJH (YS edition) [siun], HMJH (DD edition) [sun]

The zhun 諄 rhyme is generally represented as [iun]. The shu 術 rhyme is represented as [iul]. The following readings are derived from analogy: 屯 (魂 rhyme) > 窀肫迍 [tun], 君 (文 rhyme) > 窘 [kun], 樽蹲 (魂 rhyme) > 遵 [tsun]. The reading

164  Final system of 崒 [tsol] is derived from the reading of its phonetic radical 卒 [tsol], which has the fanqie spelling cangmo 倉沒 and zangmo 臧沒 in Guangyun. The reading of 蟀率 [sol] can be explained by the influence of its initial, which is retroflex [ʂ]. This initial causes the dropping of glide [j], and the place of articulation becomes lower; thus, the input form was adapted as [sol] in SK (Kōno 1979: 505). 4.11.8 Wen 文 rhyme Table 4.61 The wen 文 rhyme in EMSK 文

物

Qieyun iun

iut

EMSK un 非 敷 奉 微 見 溪 群 影 曉 云 i ii iii iv

f fʰ v ɱ k kʰ g ʔ h ɦ

on

pun 粉糞分饙奮 pun 紛忿 pun 坌𢅯墳枌濆豶鼢焚憤 mun 問璺聞文蚊紋紊 kun 君軍皸莙 kun 郡裙群 hun 訓葷纁勳薰 un 暈餫耘雲云韻紜運

ɨn

mɨn 吻ii

on 緼醖蘊iv

ul

pul 佛

ɨl pɨl 不芾弗 pɨl 拂i pɨl 咈 mɨl 物iii

kul 𨧱屈 kul 掘 ul 蔚熨鬱

HMJH, YH, IG, SG [pɨl], HD, SU, JU [pul] HMJH, HD, SU, JU [mɨn] HMJH, HG, YH, CJM, SE, IG, SG, JY, SH [mɨl], HD, SU, JU [mul] YH [on]

The wen 文 rhyme is generally represented as [un]. The readings of 緼醖 [on] are derived from the analogy of the hun 魂 rhyme characters 昷溫 [on]. The wu 物 rhyme is generally represented as [ul]. The [ɨl] reading for the wu rhyme, according to Lee (1997: 238), is the result of the unstable contrast in the feature of [labial roundness] in Old Korean. 4.11.9 Summary Table 4.62 is a summary of the mainstream representation of the zhen 臻 rhyme group in EMSK.

4.12  Dang 宕 rhyme group 4.12.1 Tang 唐 rhyme The kaikou tang 唐 rhyme is represented as [aŋ], and the hekou tang 唐 rhyme is represented as [oaŋ] in EMSK. The reading of 肓 is [haŋ] in YH. This is probably a copyist’s error, since the semantic notation for this character in YH is 膏肓 [ko

Table 4.62  Mainstream representation of the zhen 臻 rhyme group in EMSK 痕 kaikou div. 1

kaikou div. 3/3A

ən

in

Qieyun EMSK

真

ɨn

ʌn

臻

殷

kaikou div. 3B

kaikou div. 3

kaikou div. 3

in

ɪn

ɨn

in

ɨn

櫛

迄

kaikou div. 3B

kaikou div. 3

kaikou div. 3

it

ɪt

ɨl

ɨl

ɨl, əl

ɣ

in

other ɨn

labial

ən

in

質 kaikou div. 3/3A Qieyun

it

EMSK

il

labial

ɨl

il 諄

文

hekou div. 1

hekou div. 3

hekou div. 3

in

iun

iun

un

沒

術

物

hekou div. 1

hekou div. 3

hekou div. 3

it

iut

iul

ul

uon un

Qieyun EMSK

other

魂 Qieyun EMSK

ɣ

w

on

uot ul

w

ol

Table 4.63 The tang 唐 rhyme in EMSK 唐

鐸

唐

kaikou div. 1

kaikou div. 1

hekou div. 1

Qieyun

ɑŋ

ɑk

w

EMSK

aŋ

ak

oaŋ, aŋ

幫

p

paŋ 蒡榜謗

滂 並 明 見 溪

pʰ b m k kʰ

疑 影 曉

ŋ ʔ h

ɑŋ

pak 餺愽煿牔 博搏 paŋ 滂磅 pak 粕膊濼 paŋ 傍螃膀 pak 箔泊礡薄 maŋ 蟒茫忙莽芒邙 mak 幕膜漠寞莫 kaŋ 𤭛崗綱剛岡 kak 閣胳各 koaŋ 光胱廣 kaŋ 炕穅康慷 kak 恪 koaŋ 纊壙曠 haŋ 伉抗 aŋ 枊 ak 腭鍔蕚鶚諤愕 aŋ 盎 ak 堊惡 oaŋ 汪 hak 壑 hoaŋ 謊荒 haŋ 肓

鐸 hekou div. 1 w

ɑk

oak

koak 彍i槨郭 koak 籗鞹廓擴 hoak 蠖雘 hoak 癨臛 (Continued )

166  Final system Table 4.63 (Continued) 唐

鐸

唐

kaikou div. 1

kaikou div. 1

hekou div. 1

Qieyun

ɑŋ

ɑk

w

EMSK

aŋ

ak

oaŋ, aŋ

匣

ɦ

haŋ 𥮕行航吭䘕𦐄

hak 鶴貉 hoak 涸ii

端 透

t tʰ

taŋ 襠璫㼕黨讜當 tʰaŋ 帑鐋湯儻

hoaŋ hoak 幌黃皇 穫鑊 隍凰篁蝗潢蟥 璜惶煌

定

d

泥 來

n l

tʰaŋ 蕩 taŋ 塘糖堂棠螳膛 螗煻唐 naŋ 曩囊 laŋ 浪狼廊螂稂朗

精 清 從

ʦ ʦʰ ʣ

tsaŋ 葬贓臧 tsʰaŋ 鶬倉蒼 tsaŋ 欌臟藏

心

s

saŋ 顙嗓磉𤸯桑喪

ɑŋ

鐸 hekou div. 1 w

ɑk

oak

tʰak 托橐籜 飥𧘐拓託 tʰik 驝iii tʰak 鐸度 lak 諾 lak 駱落珞絡 nak 洛 tsak 作 tsʰak 錯 tsʰak 鑿 tsak 柞昨 sak 索

i HMJH [koak], HD, SU, JU [hoak] ii YH, HD [hoak], OP [hak] (popular reading [hoak]) iii HMJH (YS edition) [thik], HMJH (DD edition) [tʰak]

hoaŋ]. The kaikou duo 鐸 rhyme is generally represented as [ak] except in the case of 驝 [tʰik]. 驝 [tʰik] is represented as [tʰak] in the DD edition of HMJH. The hekou duo 鐸 rhyme is represented as [oak] in EMSK. 4.12.2 Yang 陽 rhyme Table 4.64 The yang 陽 rhyme in EMSK

非 敷 奉 微

陽

藥

kaikou div. 3

kaikou div. 3

hekou div. 3

Qieyun

iaŋ

iak

w

EMSK

iaŋ, aŋ

iak, ak

oaŋ, aŋ

f fʰ v ɱ

陽

iaŋ

paŋ 舫方昉放 paŋ 紡芳訪 paŋ 房魴防 maŋ 網輞鋩望 亡罔忘妄

藥 hekou div. 3 w

iak

oak, iak

陽

藥

陽

藥

kaikou div. 3

kaikou div. 3

hekou div. 3

hekou div. 3

Qieyun

iaŋ

iak

w

EMSK

iaŋ, aŋ

iak, ak

oaŋ, aŋ

oak, iak

見

k

kak 腳

koaŋ 誑

hoak 钁

溪

kʰ

kaŋ 襁薑疆姜韁 礓菃僵 kaŋ 羌蜣

kak 却

群 疑 影 曉

g ŋ ʔ h

koaŋ 眶筐𨴑匡恇 koaŋ 誆狂

云 來

ɦ l

知

ʈ

徹 澄 娘 精 清

ʈʰ ɖ ɳ ʦ ʦʰ

從 心

ʣ s

邪

z

莊 初 崇 生 章

tʂ tʂʰ dʐ ʂ tɕ

昌

tɕʰ

禪

dʑ

書

ɕ

日

ȵ

以

j

kaŋ 糨強 aŋ 仰 aŋ 秧鴦胦央殃怏 hiaŋ 饗鄉香響 向享 liaŋ 兩輛粮樑 蜋梁涼粱良 亮量 tiaŋ 帳張 tʰiaŋ 脹漲 tʰiaŋ 韔悵暢 tiaŋ 丈杖腸場長 iaŋ 釀 niaŋ 娘孃 tsiaŋ 醬將漿槳獎 tsʰaŋ 槍搶 tsiaŋ 鏘 tsiaŋ 匠樯薔嫱牆 siaŋ 鮝相 箱廂緗想 iaŋ 驤 siaŋ 象橡翔庠 詳祥 tsaŋ 粧裝莊壯 tsʰaŋ 瘡創 tsaŋ 狀 saŋ 牀床 saŋ 霜孀爽 tsiaŋ 掌嶂章獐障 tsʰiaŋ 彰 tsʰiaŋ 廠唱菖 娼昌敞倡 tsiaŋ 𩌬 siaŋ 嘗裳上常尙 siaŋ 觴晌賞商傷 tiaŋ 塲 hiaŋ 餉 ziaŋ 攘壤讓 siaŋ 償 iaŋ 禳蘘瓤 iaŋ 養羊陽洋楊 烊揚恙樣

hak 瘧虐 iak 約

liak 掠略 tʰiak 着 tʰiak 著 tsiak 雀爵 tsiak 鵲 tsiak 嚼 siak 削

tsiak 勺妁繳 酌灼 tsiak 綽 tsiak 杓 siak 芍 siak 鑠 ziak 箬 iak 弱若 iak 籥鑰躍 爚藥

i HMJH [iak], HD[iak] (popular reading [hoak]) SU, JU [hoak]

iaŋ

w

iak

oaŋ 枉 hoaŋ 怳貺 oaŋ 王往

iak 籰i

168  Final system The kaikou yang 陽 rhyme is represented as [iaŋ] or [aŋ]. It is generally represented as [aŋ] when appearing after labial, retroflex and velar initials and [iaŋ] elsewhere. The kaikou yang 陽 rhyme was adapted as [aŋ] when the initial is a retroflex. This can be explained by the influence of the retroflex initial, which leads to the weakness of the [j] glide. The reading of 槍搶瘡 [tsʰaŋ] is derived from the analogy of the tang 唐 rhyme character 倉 [tsʰaŋ]. The hekou yang 陽 rhyme is generally represented as [oaŋ] and as [aŋ] after labial initials. The kaikou yao 藥 rhyme is generally represented as [iak] and as [ak] when the initial is velar. The hekou yao 藥 rhyme is represented as [oak]. Below lists some hekou yao 藥 rhyme characters in other sources: 矍 [hoak](IG), 攫 [koak] (HD, OP (popular reading [hoak])), 戄 [hoak] (HD). 4.12.3 Summary Below is a summary of the mainstream representation of the dang 宕 rhyme group in EMSK. Table 4.65  Mainstream representation of the dang 宕 rhyme group in EMSK 唐

鐸

陽

藥

Qieyun EMSK

kaikou div. 1 ɑŋ aŋ

kaikou div. 1 ɑk ak

kaikou div. 3 iaŋ velar/labial/retroflex aŋ

kaikou div. 3 iak velar other ak iak

唐

鐸

陽

藥

Qieyun EMSK

hekou div. 1 w ɑŋ oaŋ

hekou div. 1 w ɑk oak

hekou div. 3 w iaŋ oaŋ

hekou div. 3 w iak oak

other iaŋ

4.13  Jiang 江 rhyme group 4.13.1 Jiang 江 rhyme Table 4.66 The jiang 江 rhyme in EMSK 江

覺

kaikou div. 2

kaikou div. 2

Qieyun

ɣ

EMSK

aŋ

幫 滂 並

p pʰ b

paŋ 邦 pʰaŋ 肨 paŋ 棒蚌

明 見

m k

paŋ 厖 kaŋ 講洚绛江扛矼豇 haŋ 港

溪 疑 影

kh ŋ ʔ

kaŋ 腔

ɔŋ

ɔk

ɣ

iaŋ

oaŋ

ak

ʌk, oak

pak 駮 pak 璞朴 pak 雹骲 鰒𩪛 mak 邈 kak 角桷覺 kak 慤 ak 嶽樂 ak 幄渥握

hoak 確i

Final system 169 江

覺

kaikou div. 2

kaikou div. 2

Qieyun

ɣ

EMSK

aŋ

iaŋ

曉 匣 知

h ɦ ʈ

haŋ 肛 haŋ 項巷缸降ii taŋ 戇

hiaŋ 𢴦

徹 澄 莊 初 崇 生

ʈʰ ɖ tʂ tʂʰ dʐ ʂ

ɔŋ

ɣ

oaŋ

tsoaŋ 樁v

taŋ 幢撞 tsʰaŋ 窗 saŋ 瀧

soaŋ 雙vi

ɔk

ak

ʌk, oak

kak 确 tʰak 卓 啄椓瘃琢斵 tʰak 踔 tʰak 濁濯擢 tsʰak 捉

hʌk 鷽iii學iv

tsʰak 鋜 sak 朔槊嗍

i YH, IG [hoak], HD, OP [kak] (popular reading [hoak]), SU, JU [kak] ii YH lists two readings: [haŋ] for the meaning “to surrender” and [kaŋ] for the meaning “to lower.” iii HMJH [hʌk], HD, SU, JU [hak] iv HMJH, HD, YH, NE, SE, IG, SH [hʌk], CJM [hʌik], JY, SU, JU [hak] v HMJH [tsoaŋ], HD, SU, JU [tsaŋ] vi  HMJH [soaŋ], YH, SG, SU, JU [saŋ], HD [s*aŋ], OP [saŋ] (standard reading [s*aŋ], popular reading [s*aŋ])

The jiang 江 rhyme is generally represented as [aŋ]. The jue 覺 rhyme is generally represented as [ak]. The reading of 𢴦 is [hiaŋ] in the YS and DD edition of HMJH, but [haŋ] in the JGG, GJG, DG edition of HMJH. The [hiaŋ] reading is probably an erroneous inscription. The readings of 樁 [tsoaŋ] and 雙 [soaŋ] are a reflection of Old Mandarin (Pak 1971; Kōno 1979). In SSTH, the notations for chong 樁 and shuang 雙 are [tsoaŋ] and [soaŋ], respectively. The readings of 確鷽 學 in MSK are listed as follows: 確 YH, IG [hoak], HD, OP [kak] (popular reading [hoak]), SU, JU [kak]; 鷽 HMJH [hʌk], HD, SU, JU [hak]; 學 HMJH, HD, YH, NE, SE, IG, SH [hʌk], CJM [hʌik], JY, SU, JU [hak]. The reading of 確 [hoak] may be derived from analogy: 穫鑊 (鐸 rhyme) > 確 [hoak]. The rhyming pattern of poems shows that the jiang 江 rhyme underwent a shift from [ɔŋ] to [ɑŋ] in the Tang Dynasty (Li 2000: 133). The reading of 鷽學 [hʌk] is a reflection of the Chinese of the Southern and Northern Dynasties, in which the phonetic value of the jiang 江 rhyme was close to [ɔŋ] (Kōno 1979: 457; Lee 1997: 148; Shin 2015: 205). 4.13.2 Summary Below summarizes of the mainstream representation of the jiang 江 rhyme group in EMSK. Table 4.67  Mainstream representation of the jiang 江 rhyme group in EMSK

Qieyun EMSK

江

覺

kaikou div. 2 ɣ ɔŋ aŋ

kaikou div. 2 ɣ ɔk ʌk

ak

170  Final system

4.14  Zeng 曾 rhyme group 4.14.1 Deng 登 rhyme Table 4.68 The deng 登 rhyme in EMSK 登

德

登

德

kaikou div. 1

kaikou div. 1

hekou div. 1

hekou div. 1

ək

w

Qieyun əŋ EMSK ɨŋ 幫 p 滂 pʰ 並 b 明 見 溪 曉 匣 端

m k kʰ h ɦ t

透 tʰ 定 d 泥 n 來 l 精 ts 從 ʣ 心 s i ii iii iv v vi vii

ʌŋ

ɨk

pɨŋ 崩

pɨk 北

pɨŋ 朋堋 棚

pɨk 踣

kɨŋ 肯 tɨŋ 鐙凳 橙燈 登等 tɨŋ 藤騰 籐 nɨŋ 能 lɨŋ 稜 tsɨŋ 罾矰 增憎 tsɨŋ 贈 tsʰɨŋ 層 sɨŋ 僧

hʌŋ 亘 hʌŋ 恒

ək

ok, ʌk, ʌik

əŋ

uŋ

oŋ, oiŋ

ək

uk, ok

pok 蔔i

mɨk 墨黙 kɨk 克 hɨk 黑 tɨk 得

w

kʌk 刻 tək 德v

koiŋ 肱ii kuk 國

iii

hʌik 劾

huŋ 薨iv

hoŋ 弘

hok 惑

tʰɨk 慝 tʰɨk 特 lɨk 肋勒鳓 tsʰɨk 則 tsək 賊vi sʌik 塞vii

HMJH, HD [pok] HMJH, SE, SG, HD, SU, JU [koiŋ] HMJH, YH, CJM, HD [kʌk], SU, JU [kʌik] HMJH, SG [huŋ] HMJH, HG, YH, CJM, NE, SE, IG, SG, JY, SH [tək] HMJH, YH, CJM, SE, SG, HD [tsək], OP [tsɨk] (standard reading [tsək]), SU, JU [tsɨk] HMJH, YH, CJM, NE, SE, IG, SG, JY [sʌik]

The kaikou deng 登 rhyme is generally represented as [ɨŋ] or [ʌŋ]. The [ʌŋ] reading is seen in 亘恒 [hʌŋ]. The [ɨŋ]/[ʌŋ] reading of the deng 登 rhyme is parallel to the reading of [ɨn]/[ʌn] in the hen 痕 rhyme. Due to the influence of the coda [ŋ], the place of articulation for the nuclear vowel in the deng 登 rhyme is marginally higher than [ə]; thus, it was generally adapted as [ɨ] in SK. As Kōno (1979) notes, the [ʌ] vowel in Middle Korean was [ə] in Old Korean. Therefore, the 恒 [həŋ] in the source input was adapted in Old Korean as [həŋ] first, and later

Final system 171 became [hʌŋ]. The reading of 亘 [hʌŋ] was probably derived from analogy of the reading of 恒 [hʌŋ] because the initial of 亘 [hʌŋ] is the jian 見 initial [k], which is not expected to be represented with [h]. The hekou deng 登 rhyme is generally represented as [uŋ] or [oŋ]. The reading of 肱 is [koiŋ] in HMJH, SE, SG, HD, SU and JU, showing that the [koiŋ] reading was widely accepted in MSK. The kaikou de 德 rhyme is generally represented as [ɨk] or [ʌk]. It is represented as [ək] in the following two characters: 德 [tək] and 賊 [tsək]. These two characters, which are frequently used characters, probably preserve the reading in the adapted form and are not affected by the sound change [ə] > [ʌ] in ancient Korean. The reading of 塞 [sʌik] probably adopted the [ʌi] from the reading of 塞 [sʌi], and the reading of 劾 [hʌik] probably adopted the [ʌi] from the reading of 劾 [hʌi] (Ito 2007: 207). The reading of 蔔 [pok] is probably derived from the analogy of wu 屋 rhyme characters 福匐 [pok]. The hekou de 德 rhyme is represented as [uk] or [ok] in EMSK. 4.14.2 Zheng 蒸 rhyme Table 4.69 The zheng 蒸 rhyme in EMSK 蒸

職

職

kaikou div. 3

kaikou div. 3

hekou div. 3

ɨk

w

Qieyun ɨŋ EMSK ɨŋ 幫 p 滂 pʰ 並 b 見 群 疑 影

k g ŋ ʔ

曉 h 云 ɦ 來 l 知 徹 澄 娘 精 從 心 莊 初 生

ʈ ʈʰ ɖ ɳ ʦ ʣ s tʂ tʂʰ ʂ

iŋ

ɨk

ik

piŋ 冰i

kɨk 棘襋 kɨk 極

ɨŋ 凝 ɨŋ 鷹膺應

ək 億vi臆 憶抑

hɨŋ 興 lɨŋ 綾菱陵䔖凌 tɨŋ 澄 tsɨŋ 甑 tsɨŋ 繒

niək 力 tiŋ 徵viii tiŋ 懲 tsɨk 鯽即 tsʰɨk 仄側 tsʰɨk 廁测惻

iək

piək 逼ii piək 堛iii pʰip 煏iv pʰiak 愎v

piŋ 憑 kɨŋ 矜

iək, ʌik, ək, ip

ɨk

tʰik 鶒敕飭 tik 直 lik 匿 tsik 稷 sik 息熄

iək 閾 iək 域vii

tʰiək 陟

tsʰʌik 昃ix sʌik 穑x嗇 色xi (Continued )

172  Final system Table 4.69 (Continued) 蒸

職

職

kaikou div. 3

kaikou div. 3

hekou div. 3

ɨk

w

Qieyun ɨŋ EMSK ɨŋ 章 昌 禪 書 船 日 以 i ii iii iv v vi vii viii ix x xi xii

tɕ tɕʰ dʑ ɕ ʑ ȵ j

iŋ

tsɨŋ 證烝蒸拯 tsʰɨŋ 秤 tsʰiŋ 稱 sɨŋ 承 sɨŋ 勝滕升 sɨŋ 乘繩 iŋ 剩 iŋ 仍 sɨŋ 蠅 iŋ 孕xii媵 賸

ɨk

ik

iək, ʌik, ək, ip

ɨk

iək

tsik 織職 sik 殖植寔 sik 飾式拭識 sik 食蝕 ik 翼弋翌杙

HMJH, YH, IG, SG [piŋ] In addition to [piək], YH also lists the reading [pʰip]. HMJH [piək], HD [phik], OP [piək] (standard reading [phik]) HMJH [pʰip], HD [phik], OP [piək] (standard reading [phik]), SU, JU [piək] YH [phiak], HD [phik] (popular reading [phiak]), OP [piək] (popular reading [phiak]) HMJH, YH, SE, IG, SG [ək] HMJH, YH, SG [iək] HMJH, YH, NE, SE [tiŋ], JY, HD, SU, JU [tsiŋ] HMJH, CJM, SE [tsʰʌik], IG [tsʰʌik]/[tsʰɨk], HD, SU, JU [tsʰɨk] HMJH, CJM, YH, SE, SG [sʌik] YH, SE [sʌik] HMJH, YH, SE, IG [iŋ]

The zheng 蒸 rhyme is generally represented as [ɨŋ]/[iŋ]. The mix of [ɨ] and [i] after [ts]/[tsh]/[s] initial is seen in Middle SK, such as 瑟 [sɨl] > [sil], 什 [sɨp] > [sip], 叱 [tsɨl] > [tsil], 輯[tsɨp] > [tsip], 則 [tshɨp] > [tship], 襯齔 [tshɨn] > [tshin] (Nam 1973: 125). The kaikou zhi 職 rhyme is generally represented as [ɨk] or [ik] following place of articulation. It is represented as [ik] for retroflex, alveolar or palatal initial, and [ɨk] elsewhere. The non-mainstream representation of the kaikou zhi 職 rhyme includes [iək], [ʌik] and [ək], which Pak (1971) and Kōno (1979) regard as borrowings from Chinese at different stages. However, the [ʌik] reading is only seen in the zhuang 莊 series initials. It is unlikely that the borrowing of Chinese pronunciation was based on a particular criterion on initials. When the kaikou zhi 職 rhyme [ɨk] occurs after the initial group zhuang 莊, which are retroflexes, the place of articulation becomes lower and moves back, and sounds like [ək]; thus, it was represented as [ək] in Old Korean, and became [ʌk] in Middle Korean. The [i] in [ʌik] was probably generated due to the influence of the soft palatal [k] coda. The input form for the zheng 蒸 rhyme may include a weak [i] before the nuclear vowel [ɨ]. Since there was probably no [iɨk] sound in Old Korean, the [iɨk] in the source input was adapted as [iək] in 堛 [piək], 力 [niək] and 陟 [tʰiək]. The kaikou zhi 職 rhyme is represented as [ək] after the ying 影 initial [ʔ], such as 億臆憶抑 [ək]. The glottal [ʔ] in the source form lowers the place of articulation for the vowel, and, thus, the zhi 職 rhyme was adapted as

Final system 173 [ək]. The reading of 煏 [pʰip] is expected to be represented as [pʰik]. The coda [k] was assimilated as [p] due to the influence of the labial initial (Ito 2007: 213). The hekou zhi 職 rhyme character is represented as [iək] in EMSK. 4.14.3 Summary Below summarizes the mainstream representation of the zeng 曾 rhyme group in EMSK. Table 4.70  Mainstream representation of the zeng 曾 rhyme group in EMSK

Qieyun EMSK

Qieyun EMSK

登

德

登

德

kaikou div. 1

kaikou div. 1

hekou div. 1

əŋ ɨŋ

ək ɨk

hekou div. 1 w əŋ uŋ

蒸

職

職

kaikou div. 3 ɨŋ ɨŋ

kaikou div. 3 ɨk ɨk

w

iŋ

ək

w ək uk

hekou div. 3 ɨk iək

ik

4.15  Geng 梗 rhyme group 4.15.1 Geng 庚 rhyme (div. 2) Table 4.71 The geng 庚 rhyme (div. 2) in EMSK 庚

陌

庚

kaikou div. 2

kaikou div. 2

hekou div. 2 hekou div. 2

Qieyun ɣæŋ EMSK ʌiŋ 幫 p 滂 pʰ

pʰʌiŋ 烹

並 b 明 m 見 k 溪 群 疑 影 曉 匣

kʰ g ŋ ʔ h ɦ

mʌiŋ 猛盲 虻孟 kʌiŋ 羹更秔 kʌiŋ 坑

æk

ɣ

iəŋ

ʌik

wɣ

æŋ

iək, ak oiŋ

pʌik 伯百 pak 迫 栢 pʌik 魄珀 拍 pʌik 帛舶 白 mʌik 陌 貊 kiəŋ 梗i綆莄 kak 骼iii koiŋ 觥 粳ii鶊鯁 kiək 格 kʌik 客

陌

wɣ

æk

oik

koik 虢

ʌik 額 hʌiŋ 杏行 胻桁

hiəŋ 亨 hiəŋ 衡iv

hiək 赫

hoiŋ 橫 (Continued )

174  Final system Table 4.71 (Continued) 庚

陌

庚

kaikou div. 2

kaikou div. 2

hekou div. 2 hekou div. 2

Qieyun ɣæŋ

云 來 徹 澄

EMSK ʌiŋ

iəŋ

ɦ l ʈʰ ɖ

tiəŋ 鋥v

莊 tʂ 初 tʂh 生 ʂ i ii iii iv v

æk

ɣ

lʌiŋ 冷 tʰʌiŋ 撐

tʰʌiŋ 鐺 sʌiŋ 生甥笙 鉎鼪省眚

ʌik

wɣ

æŋ

iək, ak oiŋ

陌

wɣ

æk

oik

tʰʌik 坼 tʰʌik 宅澤擇 tsʰʌik 笮窄 tsʰʌik 柵

HMJH, YH, SG [kiəŋ] HMJH, HD [kiəŋ], SU, JU [kʌiŋ] HMJH [kak], YH, SU [kiək], HD [kiək] (popular reading [hʌik]) HMJH, YH, CJM, SE, SG [hiəŋ] HMJH [tiəŋ], HD, SU, JU [tsiəŋ]

The kaikou geng 庚 rhyme (div. 2) is generally represented as [ʌiŋ]. Pak (1971: 169) claims that the [i] in [ʌiŋ] results from the fact that the -ng coda in the geng 梗 rhyme group in the source input contains a palatal [i], namely [ing]. The [iəŋ] reading in some characters probably results from analogy, such as 呈 (清 rhyme) > 鋥 [tiəŋ]. Kōno (1979) notes that the reading of [iəŋ] after velar/guttural initials cannot be explained by analogy and views this as a reflection of the [i̯ äŋ] form of the geng 庚 rhyme in Old Mandarin. However, the two characters 羹粳 are listed as homophonous characters in ZYYY, while they are represented as [kʌiŋ] and [kiəŋ], respectively, in SK. The insertion of [i] after velar/guttural initials in SK is probably a reflection of the [ɣ] glide in the source language. The [ɣ] glide after velar/guttural initials is more salient and sounds like [j], and therefore was adapted as [iəŋ]. This is supported by a similar sound change in Chinese where some of the velar/guttural initial characters, such as 杏行, are pronounced as [ɕiŋ] in Mandarin. The hekou geng 庚 rhyme (div. 2) is represented as [oiŋ]. The kaikou mo 陌 rhyme (div. 2) is generally represented as [ʌik]. The reading of 骼 [kak] is derived from the analogy of its phonetic component 各 [kak], which is a duo 鐸 rhyme character. The reading of 迫 [pak] may derived from tang 唐 rhyme character 粕 [pak]. The hekou mo 陌 rhyme (div. 2) is represented as [oik] in EMSK. 4.15.2 Geng 庚 rhyme (div. 3) The kaikou 庚 rhyme (div. 3) is represented as [iəŋ], except in the case of 盟 [mʌiŋ]. There are two fanqie spellings for 盟 [mʌiŋ] in Jiyun. The first reading is meibing 眉兵, a third division geng 庚 rhyme with a ping tone, while the second reading is mogeng 莫更, a second division geng 庚 rhyme with a qu tone. The [mʌiŋ] reading of 盟 is consistent with the mogeng 莫更 fanqie spelling. The

Final system 175 Table 4.72 The geng 庚 rhyme (div. 3) in EMSK 庚

陌

庚

kaikou div. 3

kaikou div. 3

hekou div. 3

iæk

w

Qieyun iæŋ EMSK iəŋ 幫 並

p b

明 見

m k

溪 群 疑 影 曉 云

kʰ g ŋ ʔ h ɦ

i ii iii iv

ʌiŋ

piəŋ 柄兵炳秉丙 piəŋ 病評 pʰiəŋ 坪枰平 miəŋ 皿命明鳴 kiəŋ 景境敬鏡京蟼驚警竟 hiəŋ 荊 kiəŋ 卿慶 kiəŋ 鯨檠黥𢐧擎競 iəŋ 迎 iəŋ 影映英

iək

ɨk

iæŋ

iəŋ

piək 碧 mʌiŋ 盟i

kɨk 戟ii kiək 綌 iək 逆

kɨk 隙iii kɨk 屐iv劇 hiəŋ 兄 iəŋ 泳蝾榮 永嶸詠咏 hiəŋ 瑩

HMJH, YH, CJM, SG, HD [mʌiŋ], OP [miəŋ] (standard reading [mʌiŋ]), SU, JU [miəŋ] HMJH, YH [kɨk] HMJH, YH [kɨk] HMJH, YH [kɨk]

hekou 庚 rhyme (div. 3) is represented as [iəŋ]. The kaikou mo 陌 rhyme (div. 3) is represented as [iək] or [ɨk]. Given that the reading of 戟 is [ki] in NGD and PTS (Hu 1963: 190), this reading is not borrowed from Old Mandarin. The [ɨk] reading is seen in the zhi 職 rhyme, such as 極棘襋 [kɨk]. In QYZZT, the zhi 職 rhyme characters 極隙 are placed in the same line. The [k] coda in [kɨk] shows that it is not borrowed from Old Mandarin, in which the [k] coda has lost. The [ɨk] reading of the mo 陌 rhyme is consistent with Late Middle Chinese, as shown in QYZZT. 4.15.3 Geng 耕 rhyme Table 4.73 The geng 耕 rhyme in EMSK 耕

麥

耕

kaikou div. 2

kaikou div. 2

hekou div. 2 hekou div. 2

Qieyun ɣɛŋ EMSK iəŋ 幫 p 明 m 見 k

ɛk

ɣ

ʌiŋ

piəŋ 迸 miəŋ mʌiŋ 黽iii 萌氓甍 kiəŋ 耕 耿iv

ɨŋ pɨŋ 繃i䙀ii

ʌik mʌik 脈麥

wɣ

iək

ɛŋ

oiŋ

麥

wɣ

ɛk

oik, uk

piək 薜 kiək 膈𥴩隔 hiək 革v

koik 馘 kuk 膕蟈vi (Continued )

176  Final system Table 4.73 (Continued) 耕

麥

耕

kaikou div. 2

kaikou div. 2

hekou div. 2 hekou div. 2

Qieyun ɣɛŋ

溪 疑 影 匣

ɛk

ɣ

EMSK iəŋ

ʌiŋ

kʰ ŋ ʔ ɦ

kʌiŋ 鏗

知 ʈ 莊 tʂ 初 tʂʰ 崇 dʐ

iəŋ 甖vii ʌiŋ 鸚鶯櫻 kʌiŋ 莖 hʌiŋ 幸倖 tsʌiŋ 箏諍爭 tsʌiŋ 崢

ɨŋ kiəŋ硬

ʌik

wɣ

iək

ɛŋ

oiŋ

ʌik 軛厄 hoiŋ 泓 hʌik 覈核 kiək 翮viii koiŋ 紘宏

tsʰʌik 謮 責 tsʰʌik 策 冊

麥

wɣ

ɛk

oik, uk

hoik 獲畫

tiək 謫ix摘 tsiək 幘x

i HMJH, HD [pɨŋ], SU, JU [phʌiŋ] ii HMJH [pɨŋ], HD, SU, JU [phʌiŋ] iii HMJH [miəŋ], HD, SU, JU [mʌiŋ] iv YH, IG, SG [kiəŋ] v  HMJH, YH, SE, IG, SG, JY, SH [hiək], HD, OP [kiək] (popular reading [hiək]), SU, JU [kiək] vi HMJH [kuk], HD [kuik], OP [koik] (standard reading [kuik]), SU, JU [koik] vii HMJH [iəŋ], HD, SU, JU [ʌiŋ] viii HMJH [kiək], HD, SU, JU [hʌik] ix HMJH, YH [tiək], HD, SU, JU [tsiək] x HMJH, YH [tsiək], HD, SU, JU [tshiək], OP [tshiək] (popular reading [tsiək])

The kaikou 耕 rhyme (div. 2) is generally represented as [ʌiŋ] or [iəŋ]. The [iəŋ] reading of 迸 is derived from analogy: 并 (清 rhyme) > 迸 [piəŋ]. The reading of 黽 is [mʌiŋ] in HD, SU and JU and [miəŋ] in YH. The [miəŋ] reading is consistent with the fanqie mugeng 母耿 in HWZY. As discussed in section 4.15.1, the [i] in [iəŋ] after velar/guttural initials is a reflection of the [ɣ] glide in the source input. The reading of 繃䙀 [pɨŋ] is derived from the analogy of the deng 登 rhyme characters 朋堋棚 [pɨŋ]. The kaikou mai 麥 rhyme is represented as [ʌik] or [iək]. The reading [iək] in the following characters is derived from analogy: 辟 (昔 rhyme) > 薜 [piək], 嫡滴 (錫 rhyme) > 謫摘 [tiək], 積蹟 (昔 rhyme) > 幘 [tsiək]. The [i] in [iək] after velar/guttural initials is probably a reflection of the [ɣ] glide in the source input. The hekou mai 麥 rhyme is represented as [oik]. The reading of 膕蝈 is [kuk] in HMJH and [koik] in SU and JU. The reading of 膕蝈 [kuk] is derived from the analogy of its phonetic component 國 [kuk]. 4.15.4 Qing 清 rhyme The kaikou qing 清 rhyme (div. 3) and the hekou qing 清 rhyme (div. 3) in EMSK are generally represented as [iəŋ]. The reading of 聘 [piŋ] is a reflection of Old Mandarin as shown in SSTH: 聘 [phiŋ]. The readings of 梬騁, regardless of their initials, are both represented as [piŋ], implying that their readings are derived

Table 4.74 The qing 清 rhyme in EMSK 清

昔

清

昔

kaikou div. 3

kaikou div. 3

hekou div. 3

hekou div. 3

Qieyun

iᴇŋ

iᴇk

w

EMSK

iəŋ

幫

p

piəŋ 餅并

滂 並 明 見 溪 群 影

pʰ b m k kʰ g ʔ

來 知 徹

l ʈ ʈʰ

澄

ɖ

精

ʦ

清

ʦʰ

從

ʣ

心

s

邪

z

siəŋ 餳

章

tɕ

昌

tɕʰ

tsiəŋ 政鉦𢁿征 整正

禪

dʑ

書 以

ɕ j

iŋ

piŋ 聘 kiəŋ 勁 kiəŋ 輕

iək

ʌik, ik

iəŋ

w

iᴇk

iək

piək 璧襞辟 piək 闢 piək 僻癖 ik 益

liəŋ 嶺領令 tiəŋ 貞 tiəŋ 蟶檉赬偵 liəŋ 逞 tiəŋ 呈裎程

iᴇŋ

miəŋ 名 kiəŋ 頸 kiəŋ 檾傾頃 kiəŋ 瓊 iəŋ 癭瓔纓嬰縈

piŋ 騁

tsiəŋ 井精旌 tsʰiəŋ 菁睛 tsʰiəŋ 清圊凊 tsiəŋ 穽情靜 淨靖 tsʰiəŋ 晴請 siəŋ 姓性 猩省i

siəŋ 誠城盛 筬成 siəŋ 聖聲 iəŋ 贏楹盈嬴 piŋ 梬ii

tʰiək 擲𧃐 tiək 躑 tsʰiək 脊鶺蹐 tsiək 襀跡積 tsʰiək 刺 tsiək 磧 tsʰiək 瘠塉 tsiək 籍 siək 昔澙磶腊舃 惜 siək 夕席汐穸 tsʰiək 隻 tsiək 炙 tsʰiək 尺斥蚇 tsiək 赤 siək石鉐碩 siək 螫釋適 hiək 弈奕 ʌik iəŋ iək 易繹驛亦 液iii腋iv䘸 頴營穎 tʰiək 蜴 掖

iək 役疫䓈

i  YH lists two readings: [siəŋ] for the meaning “to look, to check” and [sʌiŋ] for the meaning “to reduce.” ii HMJH [piŋ], HD [iəŋ], SU, JU [iŋ], OP [iŋ] (standard reading [iəŋ]) iii HMJH, YH [ʌik] iv HMJH, YH [ʌik]

178  Final system from the analogy of 聘 [piŋ]. The kaikou xi 昔 rhyme (div. 3) and the hekou xi 昔 rhyme (div. 3) in EMSK are generally represented as [iək]. The reading of 益 is [ik] in YH and CJM. In QYZZT, 益 is placed in the same line with the zhi 職 rhyme characters 息弋. The [k] coda in [kɨk] shows that it is not borrowed from Old Mandarin, in which the [k] coda has lost. The [ɨk] reading of 益 is consistent with Late Middle Chinese as shown in QYZZT. The reading of 液腋䘸掖 [ʌik] is consistent with the second division mo 陌 rhyme and mai 麥 rhyme. As discussed above, the mix of [ʌik] and [iək] is seen in the geng 梗 rhyme group. The [ʌik] reading of 液腋䘸掖 may be a result of hyper correction. 4.15.5 Qing 青 rhyme Table 4.75 The qing 青 rhyme in EMSK 青

錫

青

kaikou div. 4

kaikou div. 4

hekou div. 4

Qieyun

eŋ

ek

EMSK

iəŋ

iək

幫 滂 並

p pʰ b

明 見 溪 疑 曉 匣 云 端

m k kʰ g ʔ ɦ ɦ t

透

tʰ

定

d

泥 來

n l

精 清 從 心

ʦ ʦʰ ʣ s

i ii iii iv v

piəŋ 瓶屏軿並 pʰiəŋ 萍 miəŋ 茗暝榠瞑冥銘溟 kiəŋ 徑經涇 kiəŋ 磬罄 hiəŋ 馨 hiəŋ 脛刑形型硎 tiəŋ 頂鼎碇飣顁丁釘疔訂 tiəŋ 汀鞓 tʰiəŋ 廳聽 tiəŋ 艇亭庭廷聤蜓定停 niəŋ 挺 niəŋ 佞 liəŋ 寧 liəŋ 鈴伶零囹 欞聆蛉鴒瓴翎靈 tsʰiəŋ 青鯖蜻 siəŋ 星鯹腥醒惺

w

ik, ʌik, ɨk

eŋ

iəŋ

piək 壁 piək 霹 piək 甓 miək 覓 kiək 墼激擊

kik 喫i ʌik 鷁ii

hiək 覡 tiək 嫡鏑靮蹢 菂𢁕滴 tʰiək 趯逖倜 惕 tiək 荻敵笛狄 糴覿迪 tʰiək滌 liək 暦癧櫪靂䍽歷 tsiək 勣績 tsʰiək 戚鏚慼 tsiək 寂 siək 錫淅蜥析 tʰiək 裼

kiəŋ 扄

hiəŋ 螢迥炯

nik 溺iii lɨk 櫟iv礫v

HMJH, YH [kik] HMJH, YH [ʌik], HD [ik], OP [iək] (standard reading [ik], popular reading [il]), SU, JU [iək] HMJH, YH [nik] HMJH, SG [lɨk], HD, SU, JU [liək] HMJH [lɨk], HD, SU, JU [liək]

Final system 179 Both the kaikou qing 青 rhyme and the hekou qing 青 rhyme are represented as [iəŋ]. The xi 錫 rhyme is generally represented as [iək]. The [ik] or [ɨk] reading, which is parallel to the [ɨk] reading of the mo 陌 rhyme, is probably a reflection of the Chinese in the Song Dynasty, in which the geng 梗 rhyme group and the zeng 曾 rhyme group has merged. In QYZZT, 喫溺 are placed in the same line with the zhi 職 rhyme characters 息弋. The reading of 鷁 [ʌik] is derived from the analogy of mai 麥 rhyme characters 搤貖 [ʌik] (Lee 1997: 198). 4.15.6 Summary Below summarizes the mainstream representation of the geng 庚 rhyme group in EMSK. Table 4.76  Mainstream representation of the geng 庚 rhyme group in EMSK 庚 Qieyun EMSK Qieyun EMSK Qieyun EMSK

kaikou div. 2 æŋ ʌiŋ iəŋ 陌 kaikou div. 2 ɣ æk ʌik iək 庚 hekou div. 2 wɣ æŋ oiŋ ɣ

庚

耕

清

青

kaikou div. 3 iæŋ iəŋ 陌 kaikou div. 3 ɣ iæk iək 庚 hekou div. 3 wɣ iæŋ iəŋ

kaikou div. 2 ɛŋ ʌiŋ iəŋ 麥 kaikou div. 2 ɣ ɛk ʌik iək 耕 hekou div. 2 wɣ ɛŋ oiŋ 麥 hekou div. 2 wɣ ɛk oik

kaikou div. 3(A) iᴇŋ iəŋ 昔 kaikou div. 3(A) iᴇk iək 清 hekou div. 3(A) w iᴇŋ iəŋ 昔 hekou div. 3 w iᴇk iək

kaikou div. 4 eŋ iəŋ 錫 kaikou div. 4 ek iək 青 hekou div. 4 w eŋ iəŋ

ɣ

Qieyun EMSK

ɣ

4.16  Tong 通 rhyme group 4.16.1 Dong 東 rhyme Table 4.77 The dong 東 rhyme in EMSK 東

屋

東

屋

hekou div. 1

hekou div. 1

hekou div. 3

hekou div. 3

Qieyun uŋ

uk

iuŋ

iuk

EMSK oŋ

ok

iuŋ, uŋ, oŋ, ioŋ iuk, uk, ok

幫 滂 並

p pʰ b

poŋ 蓬篷

明

m

moŋ 矇蠓蒙懵

pok 卜 pok 醭 pok 僕 pʰok 曝 mok 木沐

moŋ 夢i

mok 目苜牧 睦 (Continued )

Table 4.77 (Continued) 東

屋

東

屋

hekou div. 1

hekou div. 1

hekou div. 3

hekou div. 3

Qieyun uŋ

uk

iuŋ

iuk

EMSK oŋ

ok

iuŋ, uŋ, oŋ, ioŋ iuk, uk, ok

kok 谷穀轂

pʰuŋ 諷楓風防 pʰuŋ 豐 poŋ 鳳ii kuŋ 宮弓躬

pok 幅福腹蝠辐 pok 蝮 pok 服鵩復馥伏 kuk 鞠菊掬鵴

kuŋ 穹 kuŋ 窮

kuk 麴

非 敷 奉 見

f fʰ v k

溪 群 影 曉 匣

kʰ g ʔ h ɦ

云 端 透 定

ɦ t tʰ d

泥 來

n l

知

ʈ

徹 澄

ʈʰ ɖ

崇 生 精

dʐ ʂ ʦ

清 從 心 章

ʦʰ ʣ s tɕ

昌 禪 書 日 以

tɕʰ dʑ ɕ ȵ j

i ii iii iv v vi

koŋ 貢功公工蚣 工攻 koŋ 孔鞚𧌆控空

kok 哭

oŋ 塕齆甕翁螉䩺甕 ok 屋 hoŋ 𦕠 hoŋ 汞閧紅洪虹鴻 kok 縠槲斛 烘篊葒

uŋ 雄熊 toŋ 蝀凍棟涷湩東 tʰoŋ 桶痛通 tok 禿鵚 toŋ 動峒衕 tok 牘犢髑 桐瞳童銅僮洞同 櫝讟瀆韣獨讀 noŋ 齈 loŋ 籠聾櫳嚨䪊竉弄 lok 轆鹿麓漉 liuŋ 窿隆 磟簏㿍 nok 祿 tiuŋ 中 tʰiuŋ 忠衷 tiuŋ 蚛仲 tʰiuŋ 沖蟲翀 siuŋ 崇 tsoŋ 鬉鯼 tsʰoŋ 緫 tsʰoŋ 葱囪聰怱 tsʰoŋ 叢 soŋ 送

tsok 鏃 tsok 簇 tsok 族 sok 速

HMJH, YH, SE [moŋ] HMJH, YH, CJM, SE [poŋ] HMJH, HD [sioŋ], SU, JU [siuŋ] HMJH, HD [sioŋ], SU, JU [siuŋ] HMJH, HG, CJM, YH, NE, SE, IG, SG, JY [tsioŋ] HMJH, SG [tsioŋ]

sioŋ 菘iii𪀚iv tsiuŋ 眾 tsioŋ 終v螽vi tsʰiuŋ 銃茺充 ziuŋ 戎 liuŋ 融

uk 栯燠

liuk 陸六戮 tiuk 竹 tʰiuk 築 tʰiuk 蓄畜 tʰiuk 軸逐柚舳妯 tsʰiuk 縮

siuk 夙蓿肅宿 tsiuk 粥 tsʰiuk 祝 siuk 俶 siuk 熟孰淑塾 siuk 叔菽 ziuk 肉 iuk 育鬻毓

Final system 181 The dong 東 rhyme (div. 1) is represented as [oŋ]. The wu 屋 rhyme (div. 1) is represented as [ok]. The dong 東 rhyme (div. 3) is generally represented as [iuŋ], [uŋ] or [oŋ] following place of articulation. It is represented as [uŋ] for velar and glottal, [oŋ] for labial and [iuŋ] elsewhere. The wu 屋 rhyme (div. 3) is represented as [iuk], [ok] or [uk] according to the place of articulation. It is represented as [ok] for labial initials, [uk] for velar and glottal initials and [iuk] elsewhere. The [iuŋ]/[uŋ]/[oŋ] reading of the dong 東 rhyme (div. 3) is in complementary distribution, implying that they should be representation of the same layer. So is the [iuk]/[uk]/[ok] reading of the wu 屋 rhyme (div. 3). The [oŋ] reading for the dong 東 rhyme (div. 3) and the [ok] reading for the wu 屋 rhyme (div. 3) after labial initials is probably a result of syllable preference in SK. Though not yet clear about the origin of this preference, it is probably due to syllable gap in Korean or for the convenience of memorizing. 4.16.2 Dong 冬 rhyme Table 4.78 The dong 冬 rhyme in EMSK 冬

沃

hekou div. 1

hekou div. 1

Qieyun

oŋ

ok

EMSK

oŋ

ok

k kʰ ʔ ɦ t tʰ d n ts

見 溪 影 匣 端 透 定 泥 精

kok 梏 hok 酷 ok 沃鋈 kok 鵠 tok 篤督

toŋ 冬 tʰoŋ 統 toŋ 疼彤 noŋ 膿農 tsoŋ 綜 tʰoŋ 宗

tok 纛碡毒

The dong 冬 rhyme and the wo 沃 rhyme are represented as [oŋ] and [ok], respectively, in EMSK. 4.16.3 Zhong 鍾 rhyme Table 4.79 The zhong 鍾 rhyme in EMSK

非 敷

鍾

燭

hekou div. 3

hekou div. 3

Qieyun

iṷo̝ŋ

iṷo̝k

EMSK

ioŋ, oŋ, iuŋ, oaŋ

iok, ok, uk

f fʰ

poŋ 封 poŋ 捧蜂鋒烽丰峯 (Continued )

182  Final system Table 4.79 (Continued)

奉 見 溪 群 疑 影 曉 來 知 徹 澄 娘 精 清 從 心 邪 章 昌 禪 書 船 日 以

鍾

燭

hekou div. 3

hekou div. 3

Qieyun

iṷo̝ŋ

iṷo̝k

EMSK

ioŋ, oŋ, iuŋ, oaŋ

iok, ok, uk

v k kʰ g ŋ ʔ h l ʈ ʈʰ ɖ ɳ ʦ ʦʰ ʣ s z tɕ tɕʰ dʑ ɕ ʑ ȵ j

poŋ 縫逢奉 koŋ 拱恭筇供鞏 koŋ 恐 koŋ 蛬共

pok 幞

oŋ 壅癰饔雍壅擁 hiuŋ 胸凶ii loŋ 壟龍 tʰioŋ 塚 tʰioŋ 寵 tiuŋ 重 noŋ 醲 loŋ 濃 tsioŋ 蹤縱 tsioŋ 從 sioŋ 竦悚 ioŋ 聳iv soŋ 駷 sioŋ 頌誦訟松 tsioŋ 種腫踵鐘鍾 tsʰiuŋ 衝 sioŋ 舂 tsoaŋ 𪄻v ioŋ 勇涌蛹踴埇容傭 鎔墉庸踊用

kok 曲蛐暼 kuk 局i跼 ok 獄玉 uk 旭iii勖 lok 綠錄 tʰiok 䕽躅 tsiok 足 tsʰok 促 sok 粟 siok 俗續 tsʰiok 燭 tsʰiok 觸 siok 屬 sok 束 siok 贖 ziok 褥 iok 辱 iok 浴谷欲

i HMJH, YH, SG, HD, SU, JU [kuk] ii YH, HD, SU, JU [hiuŋ] iii HMJH, SG, HD, SU, JU [uk] iv The reading of 聳 is [ioŋ] for the meaning “pop up” and [sioŋ] for the meaning “be scared of” in Modern SK. v HMJH [tsoaŋ], SU, JU [sioŋ], HD, OP [sioŋ] (popular reading [tsoaŋ])

The zhong 鍾 rhyme is represented as [ioŋ], [oŋ] or [iuŋ] following place of articulation. The [iuŋ]/[uŋ]/[oŋ] reading of the zhong 鍾 rhyme (div. 3) is in complementary distribution, implying that they should be representation of the same layer. The readings of 𪄻 in EMSK are given as follows: HMJH [tsoaŋ], SU, JU [sioŋ], HD, OP [sioŋ] (popular reading [tsoaŋ]). In HMJH in the notation for the character 𪄻, it says, “𪄻 is also written as 𪁈.” The 𪄻 [tsoaŋ] reading is probably derived from the reading of 𪁈 in Old Mandarin (Lee 1990: 81; Lee 1997: 257). The zhu 燭 rhyme is generally represented as [iok] or [ok]. The readings of 局 [kuk] and 旭勖 [uk] cannot be explained by syllable preference in SK, since both [kok] and [ok] exist in Middle SK. In Table 4.79, the corresponding non-entering tone representation for 旭勖 [uk] is 胸凶 [hiuŋ]. This shows that the [u] vowel here is probably a faithful representation of the input. The [uk] and [ok] representation for the

Final system 183 zhu 燭 rhyme can be explained by different adaptation strategies. As stated in Chapter 2, Mandarin -ong [uŋ] is transcribed as [uŋ] in the Korean government transcription, but as [oŋ] in Choe-Kim (1985) and Eom (2002), indicating that the same source form may be adapted differently by different borrowers. The phonetic reconstruction of the zhu 燭 rhyme in the source input is [iṷo̝k]. Due to the influence of the medial [i] and the glide [ṷ], which are high vowels, the nuclear vowel [o̝] might be perceived as [u]. The reading of 旭勖 in the source input was probably adapted as [huk] in the very beginning and later became [uk] in a process of lenition of the initial [h]. 4.16.4 Summary Table 4.80 summarizes the mainstream representation of the tong 通 rhyme group in EMSK. Table 4.80  Mainstream representation of the tong 通 rhyme group in EMSK Div. 1 東

屋

冬

沃

Qieyun uŋ EMSK oŋ

uk ok

oŋ oŋ

ok ok

Div. 3 東

屋

鍾

燭

Qieyun iuŋ EMSK iuŋ

iuk iuk

uŋ

iṷo̝ŋ ioŋ

uk

oŋ

iṷo̝k iok

iuŋ

ok

uk

4.17 Summary Based on the above discussion, below is a summary of the mainstream representation of Middle Chinese rhymes in EMSK. 4.17.1  Kaikou first division Rhyme Group Rhyme Qieyun EMSK

果 歌 ɑ a

蟹 效 咍 泰 豪 侯 əi ɑi ɑu u ʌi ai o u

咸 覃 合 談 盍 əm əp ɑm ɑp am ap am ap

山 宕 寒 曷 唐 鐸 ɑn ɑt ɑŋ ɑk an al aŋ ak

曾 登 德 əŋ ək ɨŋ ɨk

4.17.2  Hekou first division Rhyme Group Rhyme Qieyun EMSK

Rhyme Group Rhyme Qieyun EMSK

果 遇 蟹 戈 模 灰 w ɑ uo uoi other

labial

oa o

ʌi

oi

臻 魂 uon other on

泰 w ɑi

山 桓 w ɑn other

oi

an 宕

labial un

Velar Guttural oan

末 w ɑt other al

曾

通

Velar Guttural oal

沒 uot

唐 w ɑŋ

鐸 w ɑk

登 德 w əŋ wək

東 屋 uŋ uk

冬 沃 oŋ ok

ol

oaŋ

oak

uŋ uk

oŋ ok

oŋ ok

184  Final system 4.17.3  Kaikou second division Rhyme Group Rhyme Qieyun EMSK

假

蟹

麻 æ

效

皆 佳 ɛi ɣɛ

夬 æi

ɣ

ɣ

ɣ

a

ʌi ai

ai

Rhyme Group Rhyme Qieyun EMSK

江 江 ɣ ɔŋ aŋ

咸

山

肴 咸 銜 ɣ æu ɛm ɣæm other labial io o am am

山 ɛn

ɣ

梗 覺 庚 ɣ ɔk ɣæŋ ak ʌiŋ

陌 ɣ æk ʌik

黠 ɛt

刪 æn

鎋 æt

ɣ

ɣ

ɣ

ɣ

an

al

an

al

耕 ɣ ɛŋ ʌiŋ

麥 ɣ ɛk ʌik

iəŋ

iək

4.17.4  Hekou second division Rhyme Group 假 Rhyme 麻 wɣ æ Qieyun EMSK a

蟹 皆 wɣ ɛi oi

佳 夬 wɣ ɛ wɣæi oai oai

山 山 黠 wɣ ɛn wɣɛt oan oal

刪 鎋 wɣ æn wɣæt oan oal

梗 庚 耕 wɣ æŋ wɣɛŋ oiŋ oiŋ

麥 ɛk oik

wɣ

4.17.5  Kaikou third division Rhyme Group 假 Rhyme 麻 Division Qieyun iæ EMSK

ia

Rhyme 止 Group Rhyme 之 Division Qieyun ɨ EMSK other

i

蟹 祭 3/3A 3B ɣ iᴇi iᴇi

iəi

əi

廢 iai

iəi

止 支 3/3A 3B ɣ iᴇ iᴇ other retroflex/ alveolar fricative or affricate i ʌ ɨi

脂 3/3A 3B ɣ i i other retroflex/ alveolar fricative or affricate i ʌ ɨi

效 微 ɨi velar, retroflex/ other labial guttural alveolar fricative or affricate ɨi

ʌ

ɨi

i

尤

宵 3/3A iᴇu

3B ɣ iᴇu

尤 3 ɨu other

io

io

iu

labial, guttural, retroflex fricative/ affricate u

Final system 185 Rhyme Group Rhyme Division Qieyun EMSK

Rhyme Group Rhyme Division Qieyun EMSK

Rhyme Group Rhyme Division Qieyun EMSK

尤 幽

咸 鹽 3/3A iᴇm iəm

iu iu

葉 3/3A iᴇp iəp

3B ɣ iᴇm iəm

深

3B ɣ iᴇp iəp

嚴

業

iam əm

iap əp

山

侵 緝 仙 3/3A 3B 3/3A 3B 3/3A ɣ ɣ im ip ip iᴇn im other retroflex fricative/ affricate im ʌm ɨm ip ɨp iən

薛 元 月 3B 3/3A 3B ɣ ɣ iᴇn iᴇt iᴇt iɤn iɤt other labial ən

iən

臻

iəl

əl

ən

əl

al

宕

真(臻) 質(櫛) 殷 3/3A 3B 3/3A 3B ɣ ɣ in it it ɨn in other labial other labial in

ɨn

Rhyme Group Rhyme Division

曾

Qieyun EMSK

ɨŋ ɨŋ

in

il

ɨl

il

ɨn

迄 陽

藥

ɨt iaŋ iak other labial, other velar velar, guttural ɨl iaŋ aŋ iak ak

梗

蒸

職 ɨk ɨk

iŋ

庚

陌

清 3/3A

iæŋ iəŋ

ɣ

iæk iək

iᴇŋ iəŋ

ɣ

ik

3B ɣ

iᴇŋ

昔 3/3A iᴇk iək

3B iᴇk

ɣ

4.17.6  Hekou third division Rhyme Group Rhyme Division Qieyun EMSK

果

遇

戈

魚

w

iɑ

oa

iɤ other iə

蟹 虞 velar guttural ə

retroflex fricative/ affricate o

iṷo̝ other iu

labial, velar guttural u

祭 3/3A w iᴇi

iəi

3B wɣ iᴇi

廢 w

iai

iəi

186  Final system Rhyme Group Rhyme Division Qieyun EMSK

止

咸

支 3/3A w iᴇ iu

Rhyme Group Rhyme Division Qieyun EMSK

Rhyme Group Rhyme Qieyun EMSK

Rhyme 梗 Group Rhyme 庚 Division Qieyun wɣiæŋ EMSK iəŋ

3B wɣ iᴇ uəi/ui

脂 3/3A w i iu

3B wɣ i uəi/ui

微

凡

乏

ɨi uəi/ui

ṷiam əm

ṷiap əp

w

山 仙 3/3A w iᴇn iən

薛 3/3A w iᴇt iəl

3B wɣ iᴇn uən

3B wɣ iᴇt iəl

臻

元

月

w iɤn uən

w

宕

諄 in iun

術 w it iul

w

文 iun un

物 iut ul

iɤt uəl

曾

陽 w iaŋ oaŋ

藥 w iak oak

職 w ɨk iək

通 陌 wɣ

-

iæk

清 昔 東 屋 鍾 燭 3/3A 3B 3/3A 3B w iᴇŋ wɣiᴇŋ wiᴇk wɣiᴇk iuŋ iuk iṷo̝ŋ iṷo̝k iəŋ iək iuŋ/uŋ iuk/ uk ioŋ/ oŋ iok/ok

4.17.7  Kaikou fourth division Rhyme Group Rhyme Qieyun EMSK

蟹

咸

齊 ei iəi

添 em iəm

山 帖 ep iəp

先 en iən

梗 屑 et iəl

青 eŋ iəŋ

錫 ek iək

4.17.8  Hekou fourth division Rhyme Group Rhyme Qieyun EMSK

蟹 齊 w ei iəi

山 先 w en iən

屑 w et iəl

梗 青 w eŋ iəŋ

Final system 187 As the above tables show, the rhymes in Middle Sino-Korean have the following features. First, the distinction in chongniu “rime doublets” is preserved in Middle SK. The contrast between 3A and 3B is significant, except for labial initials. For example, 3B division is represented with a [ɨ] sound in kaikou zhi 支 rhyme (3A [i], 3B [ɨi]), kaikou zhi 脂 rhyme (3A [i], 3B [ɨi]), kaikou zhen 真 rhyme (3A [in], 3B [ɨn]) and kaikou qin 侵 rhyme (3A [im], 3B [ɨm]). The [i] glide in 3A division is dropped in 3B in kaikou ji 祭 rhyme (3A [iəi], 3B [əi]), kaikou xian 仙 rhyme (3A [iən], 3B [ən]), kaikou yan 鹽 rhyme (3A [iən], 3B [ən]), hekou zhi 支 rhyme (3A [iu], 3B [uəi]), hekou zhi 脂 rhyme (3A [iu], 3B [uəi]), hekou ji 祭 rhyme (3A [iəi], 3B [ui]) and hekou xian 仙 rhyme (3A [iən], 3B [uən]). Zhou (1989: 107) pointed out that “the 3A division characters and the 3B division characters are not used as the lower characters for fanqie mutually in Jingdian shiwen 經典釋文 (c. 600), XYYY (c.650), HLYY (810), Zhu Ao’s 朱翱 fanqie annotation and Jiyun 集 韻 (early eleventh century), indicating that the contrast between div. 3A and div. 3B existed in Chinese during this long period.” The preservation of chongniu in Middle SK implies that the main layer of SK was borrowed from Middle Chinese. Second, given the phonotactic constraints in Korean, the glides in Middle Chinese are always simplified in the process of adaptation, such as hekou yang 陽 rhyme [wiaŋ][oaŋ]. Second division glide [ɣ] is usually lost in the adapted form, but occurs as [i] in some cases, such as the [io] reading of the yao 肴 rhyme. Third division glide [j] is generally adapted as [i], but it is often dropped when the initial is labial, velar or guttural. Third, other features in the rhymes of Middle SK include keeping the distinction between some double rhymes, keeping the distinction between /p/, /t/ (represented as /l/) and /k/ in the coda position, and fourth division rhymes containing a [j] glide and representing the zhi 止 rhyme group (jing 精 initial group and zhuang 莊 initial group) with a [ʌ] sound. These features can provide us with valuable information about the origin of Modern SK, and a detailed discussion on these features will be provided in Chapter 6.

5 Tonal system of Middle Sino-Korean

This chapter compares the tonal system in Middle Chinese and Middle SinoKorean. There are four tones in Qieyun 切韻, namely ping 平, shang 上, qu 去 and ru 入. A general accepted description on these four tones can be expressed as follows: ping tone is a level tone, shang tone a rising tone, qu tone a falling tone and ru tone an abrupt tone (Pan and Zhang 2013). There are three major changes that occurred in the Chinese tonal system during the historical period between Middle Chinese and Old Mandarin. First, the ping tone was split into the yin 阴 register (high register), where the initial is voiceless, and the yang 陽 register (low register), where the initial is voiced. Second, the shang tone in voiced initial syllables changed to the qu tone. Third, the -p, -t, -k coda in the ru tone syllables eventually vanished, and the ru tone merged into the ping, shang and qu tones.

5.1  Tones in HMJH (Yesan mungo edition) In HMJH, Choe states, “the pitch of sound is indicated through the number of dots drawn beside the characters. There is no dot for ping tone, two dots for shang tone, and one dot for qu tone or ru tone.”1 For the convenience of transcription, this study uses L (as in Level) to refer to the ping tone, R (as in Rising) for the shang tone and H (as in High) for the qu/ru tones. Below is an example of our transcription of the tones in HMJH. Table 5.1  Tone transcription convention in HMJH Tone

Character

Representation in HMJH

Transcription

Phonetic description

ping shang qu ru

天 宇 地 節

no dot two dots one dot one dot

L R H H

level rising high high

The following table lists the representation of MC tones in HMJH (YS edition). It is noteworthy that for some characters, such as 關 [koan], 蛊 [ko] and 聵 [huəi],

Tonal system 189 two tone notations can be found. In such cases, only the first tone notation was included in the following chart. Table 5.2  Representation of MC tones in HMJH (YS edition) Tone in MC ping

shang qu ru

Representation in HMJH

Number of tokens

Total number of tokens

Percentage

L R H L R H L R H L R H

1413 36 24 26 472 108 50 506 123 34 13 555

1473

95.93% 2.44% 1.63% 4.29% 77.89% 17.82% 7.36% 74.52% 18.11% 5.65% 2.16% 92.19%

606 679 602

In HMJH, the MC ping (level) tone is generally represented as the level tone (notated with no dot); the MC shang (rising) tone is represented as the rising tone (two dots), the MC qu (falling) tone, the rising tone (two dots); and, last, the MC ru (abrupt) tone, the high tone (one dot). Table 5.3  MC sources of tone representation in HMJH Representation in HMJH

MC Tone

Number of tokens

Total number of tokens

Percentage

L (no dot)

ping shang qu ru ping shang qu ru ping shang qu ru

1413 26 50 34 24 108 123 555 36 472 506 13

1523

92.78% 1.71% 3.28% 2.23% 2.96% 13.33% 15.19% 68.52% 3.51% 45.96% 49.27% 1.27%

H (one dot)

R (two dots)

810

1027

The above chart shows the following features: First, the L tone in HMJH is mainly derived from the MC ping tone (92.78%). Second, the H tone (one dot) in HMJH is mainly derived from the MC ru tone (68.52%) and qu tone (15.19%). This is consistent with Choe’s claim of representing the MC ru and qu tones as H (one dot). Third, the R tone (two dots) in HMJH is mainly derived from qu tone (49.27%) and shang tone (45.96%). This shows that the representation of the MC

190  Tonal system shang tone and qu tones have no difference in HMJH. This is consistent with the record in other documents. In the preface of DGJU it says “the shang tone and qu tone have no difference in character reading (字音則上去無別).” In SH, it says, “the shang and qu tone merged together in colloquial reading in recent time (近世 時俗之音上去相混).” All of these records point to the same phenomenon: that is, there was no distinction between the shang tone and the qu tone in fifteenth- and sixteenth-century Korean. Given that Choe represents the MC ru tone and qu tone as the same H (one dot), the MC shang, qu and ru tones are in fact represented as the same type in Korean in Choe’s time. In summary, the tonal system of SK in Choe’s time can be categorized into two types, namely, level tone and oblique tone.

5.2  Tone notation in other editions of HMJH The tone notation in different editions of HMJH is not entirely consistent. Table 5.4 lists the different tone notations given in the YS edition and the DD edition. Table 5.4  Tone notations in different editions of HMJH

夏 晷 樝 寤 磉 堗 䴲 䘸 縷 髢 布 疥 蛊 卸 度 聵 勩 七 內 焚 終

i

YS edition

DD edition

MC

ping (L) qu/ru (H) shang (R) qu/ru (H) qu/ru (H) shang (R) ping (L) ping (L) ping (L) ping (L) qu/ru (H) ping (L) qu/ru (H) (cadelle) shang (R) (parasite) qu/ru (H) qu/ru (H) ping (L) ping (L) qu/ru (H) [huəi] shang (R) [koi] ping (L) shang (R) qu/ru (H) qu/ru (H) shang (R)

shang (R) shang (R) ping (L) shang (R) shang (R) qu/ru (H) qu/ru (H) qu/ru (H) shang (R) shang (R) shang (R) shang (R) shang (R)

qu shang ping qu shang ru ru ru shang qu qu qu shang/qu

shang (R) shang (R) shang (R) shang (R) shang (R) [huəi]

shang qu qu qu qu

shang (R) qu/ru (H) shang (R) ping (L) ping (L)

qu ru qu ping ping

i In Jiyun the fanqie spelling of xia 夏 is haijia 亥駕 when the meaning is “summer”, and haiya 亥雅 when the meaning is “China.” In HMJH the meaning of xia 夏 is “summer.”

Tonal system 191 In most cases, tone notation in the DD edition is correct. However, as for the following characters 寤布卸聵內, notation in the DD edition is erroneous, but the notation in the YS edition is correct. Last, neither DD nor the YS edition is correct in the tone notation of 夏髢疥度 勩.

5.3  Reasons for erroneous tone notation in HMJH It is difficult to present a clear explanation as to each case of erroneous tone notation in HMJH; however, the analogy based on phonetic component most likely plays an important role. Below are some examples. Table 5.5  Erroneous tone notation caused by analogy

批 淹 臊 髇 煺 瀰

MC tone

HMJH

Analogy source

ping ping ping ping ping ping

shang shang shang shang shang shang

比 奄 澡 稿, 搞, 槁 腿 爾

The tone notation of some characters may be a reflection of corresponding tonal values in Old Mandarin. For example, zan 咱 has two readings in HMJH, namely, L tone [tsa] and R tone [tsam]. The R tone may be derived from the reading of zan 咱 in Old Mandarin, which is a rising yangping 陽平 tone. In addition, it is noteworthy that some time different tones may be assigned to distinguish meaning; for example, the annotation for guan 關 in HMJH is “crossroad [koan] (one dot); to close, to shut [koan] (no dot).” However, the pronunciations for these two meanings have no difference in Chinese. Only one fanqie for guan 關 can be found in Guangyun, Jiyun and Zhongyuan yinyun.

5.4  Tone representation in other documents As opposed to HMJH, which proposes to categorize the MC four tones into three types, CJM and YH clearly maintain the distinction of the four tones. In these two texts, there is no dot for ping tone and ru tone, and a dot to the left of the character for shang tone and a dot to the right for qu tone. And the ru tone can be distinguished from the ping tone by the -p, -l, -k codas. According to Synn (2006: 135– 136), accuracy in representing MC tones in YH is 99.4% for the ping tone, 92.9% for the shang tone, 94.4% for the qu tone and 99.7% for the ru tone. The mix between the shang tone and qu tone accounts for 54.1% of all mismatched cases. One can notice that there is much inconsistency in the tone notation among different eonhae works; for example, 稼 [ka] is notated with an R tone in MJ, but H tone in NE. Next, tone representation in the same text may appear variable; for

192  Tonal system example, in MJ, the tone notation for 駕 [ka] is the R tone in chapter four verse seven, but the L tone in chapter two verse 40 (Kwon 2005). In rhyme books, such as HD, SU and JU, all characters are placed according to the four tones in MC. Therefore, no tone notation is needed to specify the tones for each character in these texts. According to Lee (1963: 142), tone notation starts to become inconsistent among different documents starting in the Middle sixteenth century, becoming completely absent in documents written after the seventeenth century.

Note 1 凡字音高低皆以字傍點之有無多少為準, 平聲無點, 上聲二點, 去聲入聲皆一點。

6 Investigation into the origin of Sino-Korean

This chapter examines the origin of SK by comparing the mainstream representation in Middle SK with the sound systems in different works of Chinese, including Qieyun (601), XYYY (c. 650), JSYY (747), HLYY (810), SYCHT (c. 1050) and QYZZT (c. twelfth century). For each of the features to be compared, we will first introduce the representation in MSK and then discuss the representation in those works of Chinese.

6.1 Initials As stated in Chapter 3, the following features in the EMSK initial system can provide us a clue about the origin of SK. First, the labial-dental fei 非 series has no distinction with the bilabial bang 幫 series. Second, the yun 云 initial is separated from the xia 匣 initial. Third, the zhi 知 series has not become an affricate or fricative. Fourth, the chan 禅 initial and the chuan 船 initial belong to the same category. 6.1.1  The bang 幫 initial group and the fei 非 initial group No distinction is made between the bang 幫 (bilabial) initial group and the fei 非 (labial-dental) initial group in MSK (幫非 > [p]/[pʰ], 滂敷 > [pʰ]/[p], 並奉 > [p]/ [pʰ] and 明微 >[m]). It should be noted that this feature is not a result of adjustment in the process of adaptation, because in the case, where there were a fei 非 [f] initial in the source input, the [f] initial would had been represented systematically as one type, such as [ph] or [h]. In other words, no distinction between the bang series and the fei series in EMSK is a strong indication that there was no fei 非 series in the source language of SK. Qieyun (601) has only one series of labial initials: [p, pʰ, b, m]. This is evident in the fact that the fanqie of Qieyun do not distinguish the fei initial group from the bang initial group, such as 篇 (芳連). In the fanqie of XYYY (c. 650), bilabial and labial-dental initials were used without distinction, including the following series of examples: 諷 (不鳳), 仳 (父美), 匾 (方殄), 蜂 (匹凶), 孚 (匹于), 潘 ( 敷袁), 泛 (匹劍, 敷劍, 孚梵), 髌 (扶忍), 𨚍 (扶必), 犛 (亡交, 亡包), 憮 (莫禹), 牧 (莫祿, 亡福), 睦 (莫祿, 亡竹), 昴 (亡飽), 眇 (亡紹), 俛 (無辯), 蜜 (亡一),

194  Investigation origin 湎 (亡善) (Wang 1982a: 2). This is a strong indication that the fei series had not occurred in XYYY. In JSYY (747), according to Shao (1981: 105), the mix in the fanqie between the bang series and the fei series accounts for 16.91%. And some characters have both bang series initial characters and fei series initial characters as their initial spellers, such as 裨 (頻卑, 苻支). These facts are evidence that the fei series had not yet occurred in JSYY (747).1 In HLYY (810), there is evidence that the fei 非 initial group had already occurred, that is, the xilian 系聯2 shows that the initial spellers for the fei 非 initial correspond with the initial spellers for the fu 敷 initial (Huang 1931: 31). There is also evidence of the appearance of the fei 非 initial group in SYCHT (c.1050), where labial-dental initial characters are listed as one individual series. The fei series is also listed as an individual group in the Song Dynasty rhyme tables like SSDZ and QYZZT. 6.1.2  The yun 云 initial The yun 云 initial, also called the yu 于 or yu 羽 initial, is represented as a zeroinitial in EMSK. In Yuanben yupian 原本玉篇 (543) and Jingdian shiwen 經典 釋文 (c. 589), the xia 匣 initial and the yun 云 initial belong to the same category (Zhou 1966: 296; Wang 1982b: 154). The yun 云 initial only appears before the third division, and it is in a complementary distribution with the xia 匣 initial in Guangyun. Dong (1996: 153–154) states that “due to the influence of the [j] medial in the third division, the yun initial is in fact a palatalized xia. This difference makes the rhyme books treat the xia initial and the yun initial as two categories.” There was no distinction between the yun initial and the xia initial in XYYY in that they could be used as initial spellers for each other, such as 㨹 (于桂), 痏 (胡 軌), 佑 (胡救), 閾 (胡域) and 熊 (胡弓, 胡宮). In JSYY, the yun initial and the xia initial have two different types of initial spellers, implying that the yun initial should was independent from the xia initial. According to Huang (1931: 25–26), the yun initial and the xia initial are distinguished in HLYY. The application of the xilian technique shows that the xia initial characters 胡扈弘脛豪毫皓號何賀鶴航話會憒回 and the yun initial characters 違葦偉爰越為于宇禹運雲隕暈尤右榮 turned out to be two different groups. In SYCHT, wang 王, a yun initial character, and yin 寅, a yi 以 initial character, are listed in the same column, indicating that the readings of the yun initial and the yi 以 initial are identical. This is consistent with the arrangement in the Song Dynasty rhyme tables, such as SSDZ and QYZZT, in which the yun initial and the yi initial are listed under the same initial, namely, yu 喻. 6.1.3  The zhi 知 initial group The zhi 知 initial, che 澈 initial and deng 澄 initial are represented as [t] or [tʰ] in MSK. According to Li (1985), this is a reflection of Old Chinese, in which there was no sheshang (retroflex stop) initials. However, the zhi initial was still a stop

Investigation origin 195 sound in the Qieyun system ([ȶ] in Karlgren, Wang Li, Dong Tonghe, Li Rong, Shao Rongfen and Zhengzhang Shangfang’s reconstruction, [ʈ] in Zhou Fagao, Pulleyblank and Pan Wuyun’s reconstruction). Therefore, the zhi initial in Qieyun, regardless of being a palatal or retroflex, should also have been adapted as [t] in Korean. In other words, no distinction between shetou and sheshang in SK is expected as long as the affricativation of the zhi series has not occurred. In XYYY the duan 端 initial characters and the zhi 知 initial characters can be used as initial spellers for each other, suggesting that the zhi 知 initial had not become affricative yet. Below are some examples cited from Wang (1982a: 1–2): 湩 (都洞, 都用, 竹用), 戆 (讀絳, 竹巷), 斵 (都角), 㭬 (都角), 啄 (丁角), 摘 (都 革), 討 (恥老), 蠆 (他邁, 敕芥, 丑芥), 惕 (聽歷, 敕歷), 洟 (敕計), 撞 (徒江), 茶 (徒加, 徒迦), 瞪 (徒萌), 懟 (徒淚), 袒 (徒莧), 擢 (徒角, 徒卓), 濯 (徒角), 喃 (女 函), 襧 (女履) and 赧 (奴盞, 女盞). According to Shao (1981: 106), in JSYY there is some mix in the fanqie for duan 端 and zhi 知 (10.31%), ni 泥 and niang 娘 (12.17%), but no mix between tou 透 and che 徹 or between ding 定 and deng 澄. In HLYY, the zhi 知, che 徹 and deng 澄 initials are still independent and have not changed into affricative, while the niang 娘 initial has merged into ni 泥. Characters with the ni 泥 or niang 娘 initials can be linked through xilian (Huang 1931: 26). The evidence of affricativation of the zhi 知, che 徹 and deng 澄 initials can be found from the Song Dynasty materials. In SYCHT, the zhi 知 series is listed after the zhao 照 series, which are affricates and fricatives, implying that in Shao Yong’s time the zhi 知 series had become affricates (Zhou 1966: 198). In Zhu Xi’s 朱熹 (1130–1200) fanqie, the zhi 知 series has mixed with the zhao 照 series (Wang 1985). In QYZZT there are many traces indicating that the zhi series has merged into the zhang 章 series (Li 2000: 78). For example, zhao 肇, which is a deng 澄 initial character, is placed under the initial of chuang 牀 in QYZZT. 6.1.4  The chan 禪 initial and the chuan 船 initial Both the chan 禪 initial and the chuan 船 initial are generally represented as [s] in EMSK, as exemplified in Table 6.1. Table 6.1  Representation of the chan 禪 initial and the chuan 船 initial in HMJH

船 禪

Character

Rhyme group

Rhyme

Labial Roundness

Division

Tone

HMJH

蛇 舌 船 時 辰 塾

假 山 山 止 臻 通

麻 薛 仙 之 真 沃

unrounded unrounded rounded unrounded unrounded rounded

3 3 3 3 3 3

ping ru ping ping ping ru

sia siəl siən si sin siuk

According to Pan and Zhang (2013), the reconstructed phonetic values for the chan 禪 and chuan 船 initial are [dʑ] and [ʑ] respectively in the Qieyun system. If the chan 禪 initial in the source language were [dʑ], then it should had been

196  Investigation origin represented in the same way as the cong 從 [dz] initial, which is represented as [ts] or [tsʰ] in MSK. The table below lists the reading of some cong 從 initial characters in HMJH. The representation of the chan 禪 initial and the chuan 船 initial in MSK implies that these two initials merged in the source language that Korean borrowed from. In Yanshi jiaxun 顏氏家訓 [The family instructions of master Yan] (c. late sixth Table 6.2  The reading of some cong 從 initial characters in HMJH Character

Rhyme group

Rhyme

Labial Roundness

Division

Tone

HMJH

坐 疾 匠 瘠 晴 叢

果 臻 宕 梗 梗 通

戈 質 陽 昔 清 東

rounded unrounded unrounded unrounded unrounded rounded

1 3 3 3 3 1

shang ru qu ru ping ping

tsoa tsil tsiaŋ tsʰiək tsʰiəŋ tsʰoŋ

century), one is told that southerners pronounced shi 石 as she 射 and shi 是 as shi 舐. This shows that in southern Chinese, chan 禪 and chuan 船 had merged at that time. According to Pan (2000: 46), the chan 禪 initial and the chuan 船 initial are distinguished in rhyme books compiled by scholars in the north, such as Cao Xian’s 曹憲 (c.541–c. 645) Boya yin 博雅音, Yan Shigu’s 顏師古 (581–645) Hanshu zhu 漢書註 and Li Shan’s 李善 (630–689) Wenxuan zhu 文選注. Qieyun follows northern Chinese usage and makes a clear distinction between these two initials. In XYYY (650), these two initials employ two different types of upper characters in fanqie. According to Zhou (1948: 395), the representation of the chan initial and the chuan initial in XYYY is generally consistent with Guangyun. In JSYY (747), the chan initial and the chuan initial are distinguished in fanqie except in the case of zhe 褶. This, according to Shao (1981: 108), is because zhe 褶 had multiple readings. In HLYY (810), the chan initial and chuan initial are not distinguished, since the initial spellers for these two initials can be linked through xilian (Huang 1931: 26). In SYCHT (c.1050) the chan initial character chen 辰 and the chuan initial character chong 崇 are listed in the same column, suggesting that these two initials are identical at that time. Although chan 禪 and chuan 船 are listed as two different initials in QYZZT, the merger of these two initials can be found in the fact that many chan initial characters are placed under the chuan initial and many chuan initial characters are placed under the chan initial (Li 2000: 84–85).

6.2 Finals As noted in Chapter 4, the following features in Middle SK rhyme system can provide us a clue about the origin of Modern SK. First, chongyun (double rhymes) are distinguished in some pairs of rhymes. Second, fourth division rhymes have a -j glide. Third, the reading of the zhi 止 rhyme group is [ʌ] when the initial is from the jing 精 series or the zhuang 莊 series. Fourth, the distinction of Middle Chinese -p, -t and -k codas is preserved.

Investigation origin 197 6.2.1 Chongyun 重韻 (double rhymes) Chongyun 重韻 (double rhymes) means two or three rhymes have the same division and openness in one rhyme group. Traditionally, this concept is used only to refer to these double rhymes in the first and second division; however, it is also used to include the third division double rhymes in recent studies. In this study, the concept of chongyun is used in its broad sense. Below is a summary of these double rhymes. Table 6.3  List of double rhymes

1st division

Rhyme group

Ping tone

Shang tone

Qu tone

Ru tone

蟹

咍

海

灰

賄

覃 談 東

感 敢 董

代 泰 對 泰 勘 闞 送

合 盍 屋

冬 皆 佳

駭 蟹

咸 通 2nd division

蟹 咸 山 梗

3rd division

遇

咸 銜 山 删 庚 耕 鱼 虞

豏 欄 潸 梗 耿 語 麌

蟹 止

流 咸 山 臻

梗 通

支 脂 之 微 尤 幽 鹽 嚴 仙 元 真(臻) 殷 諄 文 庚 清 東 鍾

紙 旨 止 尾 有 黝 琰 儼 獮 阮 軫 隱 準 吻 梗 靜 腫

宋 怪 卦 夬 陷 鑑 襇 諫 映 諍 御 遇 祭 废 寘 至 志 未 宥 幼 豔 釅 線 願 震 焮 稕 問 映 勁 送 用

沃

洽 狎 黠 鎋 陌 麥

葉 業 薛 月 質 迄 術 物 陌 昔 屋 燭

198  Investigation origin Most of these double rhymes are represented as the same type in EMSK. This may result from the following potential reasons. First, the source language that SK borrowed from has no difference in these rhymes. Second, the different source forms in Chinese were adapted into the same form in Korean due to the phonotactic mismatch between Chinese and Korean. Third, two different forms in SK merged into one type due to sound change. Therefore, the fact that those rhymes are represented identically is not sufficient to provide reliable information about the borrowing time. In this study, we will focus on the following pairs of double rhymes, which are represented differently in EMSK. Table 6.4  Distinction of double rhymes in EMSK

1st division 3rd division

Rhyme group

Double Rhymes

EMSK

xie 蟹 mu 模 zhen 臻 zhen 臻 tong 通 tong 通

hai 咍 vs. tai 泰 yu 鱼 vs. yu 虞 zhen 真 vs. yin 殷 zhi 質 vs. qi 迄 dong 東 vs. zhong 鍾 wu 屋 vs. zhu 燭

[ʌi] vs. [ai] [iə]/[ə]/[o] vs. [iu]/[u] [in] vs. [ɨn] [il] vs. [ɨl]/[əl] [iuŋ] vs. [ioŋ] [iuk] vs. [iok]

6.2.1.1 Hai 咍 rhyme and tai 泰 rhyme (kaikou) The hai 咍 rhyme is distinguished from the tai 泰 rhyme (kaikou) in Qieyun. In XYYY the dai 代 rhyme characters 載代礙戴溉賚愛嘅再賽 and the tai 泰 rhyme characters 艾蓋貝害匄賴害昧沛眛 have different lower characters in fanqie. In JSYY, the dai 代 rhyme characters and the tai 泰 rhyme characters can be used to spell each other, such as 艾(五愛),3 埭(達賴) and 慨(苦蓋). Moreover, the tai 泰 rhyme character lai 賴 and the dai 代 rhyme character lai 徕 are homophones (Shao 1981: 111). In HLYY these two rhymes are identical, for the fanqie of the following characters 艾蓋太汰害大逮載戴代怠愛靉礙耐岱賚帶 can be linked through xilian. In SYCHT, the hai 咍 rhyme and the tai 泰 rhyme belong to the same category, which is reconstructed as [ai] in Zhou (1966: 645). The merging of these two rhymes is also manifested in the Song Dynasty rhyme tables. For example, in SSDZ the dai 代 rhyme characters 慨代耐載菜賽愛 and the tai 泰 rhyme characters 蓋艾帶泰害 are placed in the same line. In QYZZT one can find the hai 咍 rhyme characters 載菜賽在 and the tai 泰 rhyme characters 蓋礚艾帶 泰大奈賴害藹餀 in the same line. 6.2.1.2 Yu 魚 rhyme and yu 虞 rhyme In Qieyun yu 魚 and yu 虞 are two different rhymes. However, according to the preface of Qieyun, the distinction of yu 魚 and yu 虞 is lost in most northern Chinese dialects at that time. In Yanshi jiaxun 顏氏家訓, Yan Zhitui 顏之推 (531–591) states that “northern people pronounce shu 庶 as shu 戍, and ru 如 as ru 儒” and “many northern people pronounce ju 舉莒 as ju 矩.” The application

Investigation origin 199 of the xilian technique shows that yu 魚 and yu 虞 are two different rhymes in XYYY (Zhou 1948: 405). In JSYY (747), the yu 魚 rhyme and the yu 虞 rhyme had merged into one type, since the following characters 區俱嶇俱衢肋朐于诹 魚盱如茹渠疏菹餘余徐雎俞鬚渝𨵦朱窬殊銖 can be linked up through xilian (Shao 1981: 115). In HLYY, these two rhymes are well distinguished. According to Huang (1931: 35), the yu 魚 rhyme characters 居魚閭驢豬閭除躇猪於疏初葅 and the yu 虞 rhyme characters 雛瑜于虞蒭愉俱拘愚隅隅衢驅駒區娛廚盂逾朱 臾榆瑜珠須愉殊須俞逾侏儒殊銖 use two different types of lower characters in fanqie. In SYCHT the yu 魚 rhyme characters 魚鼠去 are used to represent both the yu 魚 and yu 虞 rhymes, suggesting that these two rhymes had merged into one type. In QYZZT the yu 魚 rhyme characters 魚豬居渠書諸除攄於虛如蜍 臚 and the yu 虞 rhyme characters 厨無鳧夫泭于貙 are listed in the same line, indicating that these two rhymes were identical at that time. 6.2.1.3 Zhen 真 rhyme and yin 殷 rhyme Qieyun makes a clear distinction of the zhen 真 rhyme and the yin 殷 rhyme. In XYYY, the zhen rhyme and the yin rhyme are distinguished in fanqie except in the following two cases: 斤(居銀) and 筋 (居銀, 居欣). Based on these two examples, Wang (1982a) claims that the zhen 真 rhyme and the yin 殷 rhyme have merged together in XYYY. However, Zhou (1948) viewed these two rhymes as two independent rhymes. Zhou states that in the annotation for jin 斤, there are two fanqie spellings, juqin 居勤 and juyin 居銀. According to the convention in XYYY, giving two fanqie spellings for one character implies there is some difference between them. Zhou argues that it is possible that besides the yin 殷 rhyme reading, jin 斤 and jin 筋 also have the zhen 真 rhyme reading. In JSYY these two rhymes had not merged into one type (Shao 1981: 116). Shao states that although there are two fanqie spellings for yin 鄞 (語斤, 牛巾), this is consistent with Guangyun, in which yin 鄞 has both the zhen 真 rhyme and the yin 殷 rhyme readings. In HLYY the zhen rhyme characters 詵臻巾銀 and the yin rhyme characters 斤 筋殷欣齦 can be linked up through xilian (Huang 1931: 38). In SYCHT the zhen rhyme, the yin rhyme and the hen 痕 rhyme are merged. The zhen 真 rhyme character chen 臣, the zhen 軫 rhyme character yin 引 and the hen 恨 rhyme character gen 艮 are used as the representative characters for the kaikou zhen 臻 rhyme group. In QYZZT the zhen 真 rhyme characters 人鄰臣申神真瞋陳紉珍巾 and the yin 殷 rhyme characters 殷勤欣 are listed in the same line, indicating that these two rhymes had merged into one type. 6.2.1.4 Zhi 質 rhyme and qi 迄 rhyme In Qieyun, zhi 質 and qi 迄 are two different rhymes. In XYYY, the zhi rhyme and the qi rhyme are distinguished in fanqie (Zhou 1948: 431–433). In JSYY, the character xi 肸 has two fanqie spellings, namely xuyi 許乙 and xuqi 許迄. Yi 乙 is a zhi 質 rhyme character, and qi 迄 is a qi 迄 rhyme character. This suggests that the

200  Investigation origin zhi rhyme and the qi rhyme are different in JSYY (Shao 1981: 116). In HLYY, qi 迄 rhyme characters never occur in fanqie. However, given that the zhen 真 rhyme and the yin 殷 rhyme have merged in HLYY, the zhi 質 rhyme and the qi 迄 rhyme should also have become the same type. In SYCHT, the zhi rhyme character ri 日 is used to represent both the zhi 質 and qi 迄 rhymes, suggesting that these two rhymes are identical. In QYZZT, these two rhymes belong to the same category because the zhi 質 rhyme characters 窒質秩叱失乙實日栗 and the qi 迄 rhyme characters 訖乞疙 are listed in the same line. 6.2.1.5 Dong 東 rhyme (div. 3) and zhong 鍾 rhyme Qieyun distinguishes the dong 東 rhyme (div. 3) and the zhong 鍾 rhyme. The application of the xilian technique shows that in XYYY the dong 東 rhyme (div. 3) characters 雄宮弓戎終螽終隆沖中蟲 and the zhong 鍾 rhyme characters 鍾 容恭兇顒匈喁龍逢封 turn out to be two different groups (Zhou 1948: 443). In JSYY, the third division dong 東 rhyme and the zhong 鍾 rhyme have two different types of lower characters in fanqie (Shao 1981: 121). In HLYY, the dong 東 (div. 3) rhyme and the zhong 鍾 rhyme are two separate rhymes. The dong 東 (div. 3) rhyme characters 弓穹隆沖中衷融嵩終螽戎 and the zhong 鍾 rhyme characters 恭邕癰凶顒鍾舂庸從从容鎔 are distinguished in fanqie (Huang 1931: 40). In SYCHT, the zhong 鍾 rhyme character long 龍, yong 甬 and yong 用 are used to represent both the dong 東 (div. 3) and zhong 鍾 rhymes. In SSDZ, the third division dong 東 rhyme characters 窮穹中忡蟲雄 are placed in the same line with the zhong 鍾 rhyme characters 恭鍾衝舂邕胸龍茸. In the No. 2 chart of QYZZT, the third division dong 東 rhyme characters 弓穹窮中忡蟲風 are placed in the same line as the zhong 鍾 rhyme characters 顒濃舂容. 6.2.1.6 Wu 屋 rhyme (div. 3) and zhu 燭 rhyme The wu 屋 rhyme (div. 3) and the zhu 燭 rhyme are two different rhymes in Qieyun. In XYYY, the wu 屋 rhyme (div. 3) and the zhu 燭 rhyme use different types of lower characters in fanqie. The lower characters of fanqie for the wu 屋 rhyme (div. 3) include 祝育煜掬鞠六逐築福 and that for the zhu 燭 rhyme include 足 欲屬蜀褥玉勖録躅 (Zhou 1948: 443). The distinction between the wu 屋 (div. 3) rhyme and the zhu 燭 rhyme remains in JSYY (Shao 1981: 121). In HLYY, the wu 屋 (div. 3) rhyme characters 縮六戮陸竹築肉鞠麴掬匊郁昱育宿 and the zhu 燭 rhyme characters 局玉獄鋦劚綠燭蓐辱欲鵒屬 can be linked into two different groups through xilian (Huang 1931). In SYCHT, the zhu 燭 rhyme character yu 玉 is listed as the corresponding hekou character for the wu 屋 (div. 3) rhyme character liu 六, implying that the wu 屋 (div. 3) rhyme and the zhu 燭 rhyme belong to the same category. In SSDZ, the wu 屋 (div. 3) rhyme characters 菊福伏郁 and the zhu 燭 rhyme characters 曲玉褥旭束 are placed in the same line. In the No. 3 chart of QYZZT, one can find the wu 屋 (div. 3) rhyme characters 竹菊叔粥伏福 目六逐郁熟 and the zhu 燭 rhyme characters 玉辱旭 in the same line.

Investigation origin 201 6.2.2  The -j glide of the fourth division rhymes There are eight fourth division rhymes in Qieyun, namely qi 齊, xiao 蕭, tian 添, tie 帖, xian 先, xie 屑, qing 青 and xi 錫. Table 6.5 lists the representation of these fourth division rhymes in EMSK. Table 6.5  Representation of fourth division rhymes in EMSK Rhyme

Qieyun

EMSK

Rhyme

Qieyun

EMSK

齊 添 先 青

ei em en eŋ

iəi iəm iən iəŋ

蕭 帖 屑 錫

eu ep et ek

io iəp iəl iək

Table 6.5 shows that all the fourth division rhymes in Qieyun include a [j] glide in EMSK, implying that in the source language that SK is based on, a [j] glide had already occurred in fourth division rhymes. In Qieyun, there is a [j] glide for the third division, while there is no [j] glide for the fourth division. In XYYY the fourth division rhyme characters and the third division rhyme characters are used to spell each other in some cases (Wang 1982a: 3–4). Specifically, the fourth division ji 祭 rhyme characters are used to spell the third division ji 霽 rhyme characters: 厲 (力計), 斃 (蒲計). The fourth division xian 霰 are used to spell the third division xian 線 rhyme characters: 戰 (之見), 顫 (之見). The fourth division xiao 蕭 are used to spell the third division xiao 宵 rhyme characters: 瞟 (普幺). The third division qing 清 are used to spell the fourth division qing 青 rhyme characters: 坰 (共營). The fourth division tian 忝 are used to spell the third division yan 琰 rhyme characters: 厭 (於簟), 黡 (於簟). According to Shao (1981: 121), JSYY uses different final spellers for the fourth division rhymes and the third division rhymes. In HLYY the third and fourth division rhymes use the same series of lower characters in fanqie. Specifically, the fourth division qi 齊 rhyme characters 麗隸帝第計繫繼濟細翳曀詣睇戾嚏㙠 can be linked up through xilian with the third division ji 祭 rhyme characters 制滯厲例藝愒罽礪曳裔 and the third division fei 廢 rhyme characters 刈乂. The merger of the fourth division rhymes and the third division rhymes is clearly established in the Song Dynasty documents. For example, in SYCHT, qi 妻, a qi 齊 rhyme character, zi 子, a zhi 止 rhyme character, and si 四, a zhi 至 rhyme character, are listed in the same column. Similarly, in QYZZT, the qi 齊 rhyme character di 低, the zhi 支 rhyme character qi 祇 and the zhi 之 rhyme character yi 頤 are placed in the same line. 6.2.3  The [ʌ] reading in the zhi 止 rhyme group The zhi 止 rhyme group (jing 精 initial group and zhuang 莊 initial group) is generally represented as [ʌ] and occasionally as [ɨi], [ʌi] or [i] in EMSK. The zhi 支 rhyme (ri 日 initial) is generally represented as [i], except for 兒 [zʌ].

202  Investigation origin Table 6.6  The [ʌ] reading of the zhi 止 rhyme group in EMSK 支

脂

之

tsʌ 姊資姿恣 諮 tsʌi 齍 tsʰʌ 次

tsʌ 子滋孳耔镃鼒兹 tsʌi 梓

tsʌ 自瓷餈 sʌ 死四肆私

tsʌ 字牸慈鶿 sʌ 笥司鸶蕬絲思伺 si 葈

sʌ 師獅螄

sʌ 姒耜祀飼寺祠詞辭嗣似 sʌ 𨧫 tsʰɨi 輜鲻 tsʌi 滓 sʌ 士事俟仕 sʌ 史使

精

ʦ

tsʌ 紫髭

清

ʦʰ

從 心

ʣ s

邪 莊 崇 生 日

z tʂ dʐ ʂ ȵ

tsʰʌ 此 tsʌ 莿雌刺 tsʌ 骴疵 tsʰi漬 sʌ 璽賜撕𪆁 sɨi 廝 sʌ 俟 sʌ 使 sɨi 簁釃 zʌ 兒 zi 爾邇

In Qieyun the fanqie speller for a zhi 止 rhyme group character with the jing 精 initial or the zhuang 莊 initial can be a zhi 止 rhyme group character with other initials. For example, the fanqie spelling for the zhi 之 rhyme character zi 子 is jili 即里, showing that zi 子 and li 里 share the same rhyme. A similar situation can be found in XYYY and JSYY. For example, in XYYY, the fanqie spelling for zi 資 is ziyi 子夷 and zisi 姊私, indicating that zi 資 and si 私 share the same rhyme with yi 夷. In JSYY, the annotation for qi 齊 says that qi 齊 and zi 資 share the same rhyme. In HLYY, the jing 精 initial characters are always used to spell the zhi 止 rhyme group characters with the jing 精 initial (Kōno 1979). The following examples are cited from Huang (1931). Table 6.7 The fanqie spellings for the zhi 止 rhyme group characters in HLYY Rhyme

支紙寘

脂旨至

精

ʦ

紫(茲此)

清

ʦʰ

資頾(子移, 紫斯, 子雌) 兹(子思),滋(子私, 咨(姊思) 子思, 子慈, 子斯) 恣(茲四, 咨肆) 子梓(茲里, 茲死)

從

ʣ

心

s

雌(此資) 刺(千利, 此恣, 此漬, 此賜, 雌四) 慈疵(自資, 自茲, 自 瓷(自咨) 辭, 疾移) 漬(資賜, 子四) 斯厮(息資, 先齎, 徙 四泗(思恣) 移)

邪

z

之止志

思(息資) 枲(思子, 思似, 思 此) 伺(先吏, 司恣, 司 次, 司利, 司寺, 司字) 辭(似茲) 似兕(辭紫, 辭姊) 寺飼(辭自, 詞字)

Investigation origin 203 Rhyme 莊 崇 生 日 章

tʂ dʐ ʂ ȵ tɕ

昌 禪

tɕʰ dʑ

書

ɕ

以

j

支紙寘

脂旨至

之止志 緇(滓師, 滓基) 滓(緇史)

爾(而紙) 支枝(止移) 紙(支爾)

師(史緇) 二餌(如志) 脂(止夷) 旨指(脂齒) 至鷙(之利)

氏是舐(時止, 時爾)

止趾(之爾, 之始, 支市) 齒(蚩耳, 鴟止) 豉時(是之) 市(時止)

翅(施至, 試異, 試豉, 尸屍(始之) 尸二) 移(以之) 夷(以之) 以已(夷止)

Table 6.7 shows that the final spellers for the jing 精 or the zhuang 莊 initial group characters tend to be the jing 精 or zhuang 莊 initial group characters. But it is necessary to point out there are many exceptions, such as 資頾 (子移), 子梓 (茲 里), 刺 (千利), 慈疵 (疾移), 斯厮 (徙移), 伺 (先吏, 司利) and 緇 (滓基). There are two possible explanations for these exceptions. First, this series of fanqie may have been copied from Qieyun or other rhyme books: For example, ci 疵 has four fanqie spellings, one of which is jiyi 疾移, which is also seen in Guangyun. Second, some characters may have more than one reading; for example, zi 頾, si 厮, zi 緇 and si 伺 have more than one fanqie in HLYY, which may be a reflection of different readings. However, it is necessary to clarify that the above statements are academic assumptions and must be subject to further investigation. In Zhu Ao’s 朱翱 fanqie for the book Shuowen jiezi xizhuan 說文解字系傳 [A collection of Shuowen jiezi] (a book written in the Five Dynasties period), the jing 精 series initial characters are only used in the fanqie spelling for the jing 精 series initial characters where the rhymes of these characters belong to the zhi 止 rhyme group. This implies that the vowel of the zhi 止 rhyme group for the jing 精 series was different from that of the zhi 止 rhyme group for other initials (cf. Wang 1982c, 1985). In SYCHT, qi 妻, zi 子, si 四 and ri 日 are placed in the same column, indicating that these four characters should share the same nuclear vowel. Therefore, the nuclear vowel of zi 子 should be -i, which is the same as the other three characters. However, Zhou (1966: 602) states that since zi 子 is listed in the first division in SYCHT, the vowel of zi 子 must have changed from [i] to [ɿ]. In the No. 18 chart of QYZZT, the following characters 兹雌慈思詞紫此死兕恣蛓自笥寺 are placed in the first division, indicating that the vowel of the zhi 止 rhyme group is not -i when the initial is from the jing 精 series. Wang (1982c) reports that based on the xieyin 叶音 information in Zhu Xi’s 朱熹 (1130–1200) Shiji zhuan 詩集傳 [Collected annotations of Shijing], the zhi 止 rhyme group (jing 精 initial group) generally became [ɿ], and the zhi 止 rhyme group (zhuang 莊 initial group) generally became [ʅ], although for some characters, there are [ʅ] and [i] two readings.

204  Investigation origin However, the zhi 止 rhyme group (zhang 章 initial group and zhi 知 initial group) is still [i]. In ZYYY, the zhi 止 rhyme group characters with initials from the jing 精 series, zhuang 莊 series, zhang 章 series, ri 日 or zhi 知 series are listed under the socalled zhisi 支思 rhyme. There is much debate regarding the phonetic value of this zhisi 支思 rhyme. Wang (1985) proposes that the phonetic value of the zhisi 支思 rhyme may be regarded as [ɿ]/[ʅ], which is identical to that in Modern Mandarin. Lu (1946) reconstructed the zhisi rhyme as [ï], claiming that [ï] is a sound between the phonetic values [ɿ] [ʅ] and values [ɨ] [ɘ]. Jin (1998) suggested that, in QYZZT, the de 德 rhyme characters 則賊塞 are used to represent the corresponding ru tone rhymes for the jing 精 series zhi 止 rhyme group, as well as zhenhen 臻痕, youhou 尤侯 and zhengdeng 蒸登 rhymes. Based on this evidence, Jin (1998) claimed that the phonetic value of the zhisi rhyme may actually be central vowel [ə]. Wang (2012) divided the development of the zhisi rhyme into two stages, claiming that the main vowel has undergone a process of centralization [i] > [ɯ] during a period from the late Tang to the Ming Dynasty. The [ʌ] reading of the zhi 止 rhyme group, according to Pak (1971), is a reflection of the [ə] in *-ï̯ ěd(-ï̯ ěi)/ *-ï̯ əd(-ï̯ əi) in a stage of Old Chinese where the coda [d] weakened into vowel [i]. However, there are two weaknesses in Pak’s hypothesis: For one, Pak’s hypothesis must deal with the circumstance that the vowel [ə] in Old Chinese is only preserved for these two initial groups. For two, in the case that the reading of the ri 日 initial character 兒 [zʌ] was borrowed from Old Chinese, the nasality of the ri 日 initial should have been retained in the phonological system of EMSK to satisfy this hypothesis. Arisaka (1936) proposed that the [ʌ] reading of the zhi 止 rhyme group was borrowed from the [ʅ]/[ɿ] in the Kaifeng 開 封 dialect in the tenth century. This study sides with Arisaka’s (1936) hypothesis. The development of the zhisi rhyme introduced above shows that it only started to occur in the jing 精 series in the late Tang Dynasty, while it had later spread to the zhuang 莊 series in the early Southern Song Dynasty as well as to the zhang 章 series and zhi 知 series in the Yuan Dynasty. Given that in SK the [ʌ] reading of the zhi 止 rhyme group is mainly seen in the jing 精 and zhuang 莊 series, it is clear that this reading should been borrowed from a variant of Chinese which was later than Tang Chinese but earlier than Yuan Chinese. 6.2.4  -l coda Among the three stop consonantal codas -p, -t and -k in MC, -p and -k are well preserved in SK; however, -t is borrowed systematically as -l in SK. The threeway distinction of the -p, -t and -k codas in MC is well kept in Qieyun, XYYY, JSYY and HLYY. However, in SYCHT, the -t and -k coda rhymes are placed in the same chart as open syllable rhymes, not following the convention of being placed after nasal coda rhymes. For example, open syllable rhyme characters duo 多, he 禾, dao 刀 and mao 毛 are placed in parallel with -t or -k coda characters she 舌, ba 八, yue 岳 and huo 霍, respectively. This fact shows that the -t and -k coda at that time had already merged into one type. However, the -p coda characters are

Investigation origin 205 placed in parallel with the -m coda, implying that the -p coda was still kept at that time. In QYZZT, the -p, -t and -k coda rhymes are always placed in parallel with open syllable rhymes. In the No.9 chart, characters de 德 (-k coda), ke 刻 (-k coda), se 瑟 (-t coda) and zhi 栉 (-t coda) are placed in the same line, showing that the -t coda and -k coda at that time had already merged into one type. Arisaka (1957) hypothesized that the -l in SK results from the weakness of -t > -l coda in the Song Dynasty. However, it should be noted that the merger of -p, -t and -k had started in the Song Dynasty. If we attribute the -l to the reading in the Song Dynasty Chinese, we would not be able to account for the systematic correspondence between the -p, -t, -k coda in MC and the -p, -l, -k in SK. Lee (1972) and Martin (1997) claimed that the -l coda in SK derived from the -r in the northwest dialect of Middle Chinese in late Tang Dynasty. Lee (1972) claimed that it is impossible to explain the appearance of -l in SK from the native Korean phonological point of view. Martin (1997: 263) claimed that “we have no reason to think that Korean ever lacked syllables ending in -t so that the syllable structure of Old Korean is not the reason Chinese -t was borrowed as -l.” Lee (1972) and Martin’s (1997) hypothesis is based on the study in Luo (1933). According to Luo (1933), the -p, -t and -k coda in the northwest dialect of MC underwent a sound change, as shown below. Table 6.8  The codas in the northwest dialect of MC Old Chinese (OC)

8th Century

9th–10th Centuries

Old Mandarin (OM)

[p] [t] [k]

[b] [d] [g]

[b] [ɹ] [g]

[β] [r] [ɣ]

However, this hypothesis is rejected by Pak (1971), Yu (1991) and Eom (2001); these scholars argue that since no evidence shows any strong relationship between the northwest dialect in the Tang Dynasty and Sino-Korean, it is not natural for Sino-Korean to borrow a sound from a dialect without such a close relationship. Also, there is a lot of evidence showing the [t] coda had been represented as [l] in SK earlier than late Tang Dynasty. For example, in Hyeseongga 彗星歌 “The song of comet” (AD 623) the word 달 [tal] “moon” had already been transcribed with Chinese character 達, which was a -t coda character in MC. Also in the SGSG -t coda character hu 忽 is used to represent the word 골 [kol] “village” in the place names in the Three Kingdom period of Korean peninsula (57 BC–AD 668). In this study, we will present a piece of evidence to show that the [t] coda in MC had already been represented as [l] in SK in the eighth century. Many Korean Buddhist monks traveled to China to learn the Buddhist classics after Buddhism was transmitted to the Korean peninsula. Some monks even made the longer journey to India. One of these monks was Hyecho 慧超 (also written as 惠超), who came from the kingdom Silla 新羅 on the Korean peninsula. Although there is no

206  Investigation origin evidence showing the exact dates regarding his birth, according to Zhang (1994), Hyecho 慧超 was probably born in AD 700 or 704. In the eleventh year in the reign of the Kaiyuan 開元 (AD 723), Hyecho travelled to India by sea, and he returned to China in AD 727. Several years after his tour of India, he studied in the Esoteric school under Vajrabodhi at the Dajianfu 大薦福 Monastery in Chang’an 長安. There, he took on a central role in the translation of Sanskrit Buddhist sutra, receiving an imperial commission to serve in the position of bishou 筆受, an official scribe and translator. In the year 780, Hyecho sojourned to the Mt. Wutai 五 台 and passed away some time later, between 780 and 783. Hyecho wrote a travelogue entitled Wang o cheonchuk guk jeon 往五天竺國 傳 [Memoir of the Pilgrimage to the Five Kingdoms of India] (hereafter Memoir) based on his travels and experiences in India between 723 and 727. The travelogue was lost for many centuries. In 1905, Paul Pelliot discovered a fragment of it in Dunhuang 敦煌. Although the manuscript only exists in fragmented pieces, Hyecho’s Memoir is thought nonetheless to be a work originally composed of three volumes. In fact, there is an entry in HLYY stating that Hyecho’s account was split up into three volumes. It seems that the extant version of the Memoir is an excerpt version of this entire manuscript. In this study, we will compare the transliteration of place names cited in Xuanzang’s 玄奘 (AD 602–664) Xiyuji 西域記 and Hyecho’s Memoir, paying special attention to the [t] coda characters. Xiyuji was written in 646 by Xuanzang. Roughly speaking, Xiyuji was compiled approximately a half century earlier than the Memoir. Table 6.9 lists some different transliterations between Xuanzang and Hyecho. The Sanskrit name for these place names are cited from Ji (1985). It should be noted that all of Hyecho’s transliterations in the following table, according to our search in the electronic version of Siku quanshu 四庫全書 (Wenyuange 文淵閣 edition), first occurred in Hyecho’s Memoir. Therefore, these transliterations are not what Hyecho adopted from other books. Table 6.9  Comparison of transliterations given by Xuanzang and Hyecho Xuanzang

Hyecho

Sanskrit

波謎羅 蘇伐剌拏瞿呾羅 那迦羅馱那 怖捍

播蜜 蘇跋那具怛羅 那揭羅馱娜 跋賀郍

pamir suvarṇagotra nagaradhana ferghana

Table 6.9 primarily provides us with three clues. The first clue concerns the first two examples in which Hyecho used only one [t] coda character, mi 蜜 or ba 跋, to reflect the [r] coda in Sanskrit; Xuanzang, however, used a [l] initial character, luo 羅 or la 剌, in addition to a [t] coda character mi 謎 or fa 伐, to reflect the [r] coda. In other words, it was necessary for Xuanzang to indicate the [r] coda through a [l] initial character, whereas it was redundant for Hyecho. The second clue concerns nagaradhana, in which the -ga is followed by a [r] initial syllable. This was transliterated with an open syllable jia 迦 by Xuanzang; however, with

Investigation origin 207 a MC [t] coda character jie 揭 by Hyecho. Third is ferghana, for which Xuanzang used the open syllable character bu 怖 to transcribe the fer, while Hyecho used the MC [t] coda character ba 跋 to do such. These three differences are clues to phonological difference between MC and Hyecho’s language; specifically, the alveolar consonant (non-nasal) coda in Hyecho’s language was closer to the [l] ([r]) sound, suggesting that the [t] coda in Middle Chinese was probably pronounced as [l] in Sino-Korean in the early eighth century. Kang (2010) explains the origin of -l coda as a customary reading, which first occurred in Buddhist temples, hypothesizing that in order to be faithful to the original -r or -l coda in Sanskrit, the -t coda characters in the Chinese translated Buddhist classics were pronounced with an -l coda by Korean monks. However, it is difficult for Kang (2010) to explain why Korean monks were only concerned about the faithfulness to the Sanskrit -l coda, rather than the faithfulness to other features of Sanskrit. Shen (2004) argued that the [t] coda became [l] through the following process: CVt > CVtV > CVrV > CVr. According to Shen (2004), Old Korean is an open-syllable language—like Japanese—not having any consonantal codas. He proposed that Korean underwent a process of the shortening of multi-syllables and the -t between two vowels changed to -l. Later the deletion of vowels in the second syllable brought about the -l coda in SK. Shen’s analysis is based on the premise that Old Korean is an open-syllable language; however, such an assumption is controversial. Although some scholars argued that Old Korean was an open-syllable language, periodizations of Old Korean are not consistent in different studies. Moreover, in the case that such proposed shortening was to happen in Old Korean, it would nevertheless be difficult to explain why this change only applied to Sino-Korean and had no influence over native Korean words, such as [pat] “field” and [nat] “sickle.” Pak (1971) claimed that the [t] coda in Middle Chinese was represented with [l] in Korean because SK avoids adapting the MC [t] as [t] for some unknown reason. However, this hypothesis is also difficult to establish given that there is no clear reason to avoid a faithful representation in SK. This study proposes that the -t coda had not occurred in Korean in the time of borrowing. As a result, the -t coda was represented as -l (which shares the same place of articulation with -t). However, this explanation needs to be established on and accounted for by continued research on the coda system of Old Korean. It is also noteworthy that the word [put] 붓 “writing brush, pen” in native Korean, which is thought to be a loanword from Old Chinese [put] 筆, seems to preserve the -t coda successfully. This may be explained through later development which was therefore unaffected. As Eom (2015: 350) states, “Old Korean did not have [t] coda and thus adapted Chinese [t] coda as [l], and the -t coda in Middle Korean occurred later than this adaptation.”

Notes 1 Shi (1983: 31) states that “the bilabial initials and labial-dental initials in Sanskrit, which were not distinguished in those old transliterations, were distinguished by Xuanzang 玄奘 (602–664).” Xuanzang was born in Goushi 缑氏, a place near Luoyang 洛陽, and lived there until his teenage years. Therefore, his transliteration should be based on the sound

208  Investigation origin system of the Luoyang dialect. Xuanzang’s transliteration shows that fei 非 series has started to occur in Luoyang in the seventh century. However, according to JSYY, there is no fei series in the Luoyang dialect in the eighth century. The inconsistency between Xiyuji and JSYY may be explained by hypothesizing that Xuanzang used the colloquial reading while He Chao used the literary reading. 2 Xilian 系聯 is a technique, first proposed by Chen Li 陳澧 (1810–1882), to categorize the upper characters of fanqie and the lower characters of fanqie. All characters that share the same series of upper characters of fanqie belong to the same initial group, and all characters that share the same series of lower characters of fanqie belong to the same rhyme group. By applying this technique, we can identify the number of initials and finals. 3 The fanqie spelling is placed in the parentheses.

7 Conclusion

The aim of the present study is to explore the origin and layers of SK, and to reconstruct the process through which the input form in Chinese was adapted into Korean. It should be noted that the hypothesis on the origin of SK provided here is to trace the origin of the SK present in the Modern period. Old Sino-Korean, as reflected in the proper names of the Three Kingdoms period that are written with Chinese characters, is probably not homogeneous with Modern SK and thus may be traced back to an origin much earlier.

7.1  Origin of Sino-Korean In Chapter 3, Chapter 4 and Chapter 5, we have identified the mainstream forms in EMSK data by examining the initial, rhyme and tones in Hunmong jahoe, Cheon ja mun and Sinjung yuhap. Some features of the mainstream forms in EMSK are due to the mismatch in the phonological system between Chinese and Korean and thus offer little help in tracing the origin. For example, the yi 疑 initial and the ying 影 initial are represented as a zero-initial in SK. The zhuang 莊 series (retroflex affricates and fricatives), the zhang 章 series (palatal affricates and fricatives), and the jing 精 series (alveolar affricates and fricatives) are all represented as the same type, namely alveolar affricates and fricatives. The glides in Middle Chinese are invariably simplified within the process of adaptation, such as hekou fei 廢 rhyme [wiai]  [iəi], hekou yuan 元 rhyme [wiɤn]  [uən] and hekou yang 陽 rhyme [wiaŋ]  [oaŋ]. The following features, however, do provide valuable clues to the origin of Modern SK.

I. Initial ① The labial-dental fei 非 series had not occurred. ② The yun 云 initial was separated from the xia 匣 initial. ③ The zhi 知 series had not yet become an affricate or fricative. ④ The chan 禪 initial and the chuan 船 initial belong to the same category.

II. rhyme ⑤ Chongyun 重韻 exists in the pair of hai 咍 and tai 泰. ⑥ Chongyun 重韻 exists in the pair of yu 鱼 and yu 虞.

210  Conclusion ⑦ ⑧ ⑨ ⑩ ⑪ ⑫

Chongyun 重韻 exists in the pair of zhen 真 and yin 殷. Chongyun 重韻 exists in the pair of zhi 質 and qi 迄. Chongyun 重韻 exists in the pair of dong 東 (div. 3) and zhong 鍾. Chongyun 重韻 exists in the pair of wu 屋 and zhu 燭. Fourth division rhymes have a -j glide. The distinction of Middle Chinese -p, -t and -k codas is preserved.

In Chapter 6, we have compared the representation of these particular initials and rhymes in EMSK with that in the following works of Chinese, including Qieyun 切韻 (601), Xuanying’s 玄應 (?–661) Yiqie jing yinyi 一切經音義, He Chao’s 何超 Jinshu yinyi 晉書音義, Huilin’s 慧琳 (737–820) Yiqie jing yinyi 一 切經音義, Shengyin changhe tu 聲音唱和圖 and Qieyun zhizhang tu 切韻指掌 圖. Table 7.1 lists the result of our comparison. The chronological and dialect background information of each book is provided for reference. In order to show whether a certain feature introduced in the manuscript is consistent with SK, we have also marked “+” (consistent) and “-” (inconsistent) in the table below. Table 7.1  Comparison of EMSK and MC Qieyun

XYYY

JSYY

HLYY

SYCHT

QYZZT

Time

601

c.650

747

810

c.1050

Area

Luoyang Nanjing + + + + + + + + +

Chang’an

Luoyang

Chang’an

+ + + + + + + + + +

+ + + + + + + +

+ + + + + +

Kaifeng Luoyang + + + + -

c.12th century Kaifeng

1 2 3 4 5 6 7 8 9 10 11 12

+ + + + -

Table 7.1 can be summarized as follows: First, the Song Dynasty materials SYCHT and QYZZT are inconsistent with SK in No. 1, Nos. 5–10 and No. 12. Second, the Qieyun system is inconsistent with SK in No. 2, No. 4 and No. 11. Third, XYYY is inconsistent with SK in No. 2 and No. 4. Fourth, HLYY is inconsistent with SK in No. 1, No. 5 and Nos. 7–10. Fifth, JSYY is inconsistent with SK in No. 4, No. 5, No. 6 and No. 11. Sixth, HLYY is consistent with SK on those points that XYYY does not agree with SK. And for those points that HLYY is inconsistent with SK, XYYY agrees with SK perfectly. This shows that the mainstream SK is a reflection of Chinese later than XYYY (c. AD 650) but earlier than HLYY (AD 810). Given that the fanqie in XYYY and HLYY is based on the pronunciation of

Conclusion 211 the Chang’an 長安 dialect, this study proposes that the matrix of Modern SK is borrowed from the Chang’an dialect between AD 650 and AD 810.

7.2  Adaptation layers of Sino-Korean It is widely believed that Sino-Korean pronunciations consist of different layers, but the number and the nature of layers is under debate (Pak 1971; Kōno 1979; Eom 1999; Synn 2006; Zhang 2008; Shin 2015). In Table 7.2, we summarize our understanding of the layers in EMSK. This study distinguishes internal layers, which result from sound change in Korean, and the layers that result from borrowing from Chinese at different stages. The four layers that originated from Chinese are named C1, C2, C3 and C4, and the internal layer that resulted from sound change in Korean is named K1. In some instances, the representation of rhymes in SK differs according to the place of articulation. The representation that occurs after a particular initial will be indicated by placing the initial in parentheses. As stated in section 7.1, the mainstream representation of SK is not the same as the phonological system in Qieyun. Where the representation of EMSK is not consistent with the phonetic reconstruction of Qieyun in Pan and Zhang (2013), we will reconstruct a phonetic value as the input form with both the trend of rhyme development in Chinese and the consistency of reconstruction taken into account. As Table 7.2 shows, EMSK includes four layers that are of Chinese origin. The mainstream representation of EMSK is C2, which is based on the pronunciation of the Chang’an 長安 dialect between AD 650 and AD 810. The C1 layer is earlier than the C2 layer and roughly corresponds to Old Chinese. It would be ideal to further divide this layer into different stages given that Old Chinese is a relatively long period, but the data we are able to access are rather limited, and this task will be left to future work. The evidence of this layer can be found in the readings of some characters which reflect the reading in OC, such as 巾 [kən] and 鷽學 [hʌk]. The C3 layer corresponds to the Chinese in the Song Dynasty. The main feature of this layer is the [ʌ] reading in the zhi 止 rhyme group. The C4 layer corresponds to Old Mandarin as shown in the Chinese transcribed in SSTH, NGD and PTS. The evidence of this layer can be found in the readings of those characters that reflect the reading in Old Mandarin, such as 騾 [lo], 刷 [soa] and 蒜 [suən]. The internal layer of EMSK results from the following sound change in Korean. First, [ʌ] merged into [a]. For example, the representative reading of the hai 咍 rhyme is [ʌi], but in some hai 咍 rhyme characters, the reading has changed to [ai]. Second, [ɨ] changed to [i], and occasionally [i] changed to [ɨ], such as the zhi 質 rhyme character 膝 [sɨl]. Third, [un] changed to [ɨn], such as the wen 文 rhyme character 吻 [mɨn]. Fourth, the effect of coarticulation caused the insertion of [i]. For example, 娵僦醜臭炊醉 are expected to be read as [tsʰiu], but are represented as [tsʰiui] in EMSK. The initial [tsʰ] causes the nuclear vowel being more fronted and rounded, and thus an [i] off-glide is inserted. Fifth, dissimilation leads to the dropping of [i] off-glide. The [i] off-glide in the following characters: 蠣 [liə], 噬

虞 咍 灰 泰

齊

廢

祭

夬

佳

合 開 合 開 合

模 魚

遇

皆

合 開 合 開 合 開 合 開 合 開 合 開

麻

假

蟹

開 合 合 開 開 合 合 合

歌 戈

歌

ɑ ɑ w iɑ ɣ a ia ʷɣa u̯ o iɤ iu̯ o̝ əi uoi ɑi w ɑi ɣ ɛi wɣ ɛi ɣ ɛ wɣ ɛ ɣ æi wɣ æi iᴇi ɣ iᴇi w iᴇi iai w iai ei w ei

3 1 1 1 1 2 2 2 2 2 2 3/3A 3B 3/3A 3 3 4 4

w

Qieyun

1 1 3 2 3 2 1 3

Table 7.2  Layers of Sino-Korean readings

iu̯ o̝ əi uoi ɑi w ɑi ɣ ɐi wɣ ɐi ɣ ɐi wɣ ɐi ɣ ai wɣ ai iei ɣ iei w iei iei w iei iei w iei

ɑ ɑ w iɑ ɣ a ia ʷɣa u̯ o iɤ w

Input

猜 [sɨi]

C1 a oa oa a ia oa o, u ə, iə, o (莊) u, iu ʌi oi, ʌi (幫) ai oi ʌi, iəi oi ʌi, iəi oai ai oai iəi əi iəi iəi uəi, iəi iəi iəi

C2

C3

話[hoa]

畫[hoa]

𡲰[tsiu]

騾 [lo]

C4

iə

iə

ai

ai

娵 [tsʰiui] ai ai

əi, iəi

左 [tsoa]

K1

咸

流

效

止

蕭 侯 尤 幽 覃 合 談 盍 咸 洽 銜 狎

豪 肴 宵

1

2

2

開 開 開 開 開

開

開

開

開 合 開 開 開

微

i i

3B 3/3A 3B 3 ɨi ɨi ɑu ɣ æu iᴇu ɣ iᴇu eu u ɨu iu əm əp ɑm ɑp ɣ ɛm ɣ ɛp ɣ æm ɣ æp w

ɨ

wɣ

w

ɣ

w

i

iᴇ iᴇ wɣ iᴇ i

3B 3/3A 3B 3/3A ɣ

iᴇ

3/3A

3 3 1 2 3/3A 3B 4 1 3 3 1

開

合

開

合

開

之

脂

支

Qieyun

i i

i i i

i

ɨi ɨi ɑu ɣ au ieu ɣ ieu ieu u ɨu iu əm əp ɑm ɑp ɣ am ɣ ap ɣ am ɣ ap w

ɨi

wɣ

w

ɣ

i

wɣ

w

ɣ

i

Input

篼 [to]

C1

am ap am ap am ap

ɨi, i uəi, ui o io, o io io io u iu, u iu ʌm

ɨi, i iu, ui uəi, ui ɨi, i

i, ɨi (精莊) ɨi, i (幫) iu uəi, ui i

C2

ʌ (精莊)

ʌ (精莊)

ʌ (精莊)

C3

C4

am ap

(Continued)

僦醜臭[tsʰiui]

醉 [tsʰiui]

i 炊 [tsʰiui]

K1

山

深

仙

鎋

刪

黠

寒 曷 桓 末 山

緝

嚴 業 添 帖 凡 乏 侵

葉

鹽

Table 7.2 (Continued)

4

3

開

合

2

開 合 開 合 開 合 開 合 開

合

3/3A 3B 3/3A 3B

2

2

2

1

合

開

3/3A 3B 3/3A 3B 1

開

開

3/3A 3B 3/3A 3B 3

開

iᴇm iᴇm iᴇp ɣ iᴇp iam iap em ep ṷiam ṷiap im ɣ im ip ɣ ip ɑn ɑt w ɑn w ɑt ɣ ɛn wɣ ɛn ɣ ɛt wɣ ɛt ɣ æn wɣ æn ɣ æt wɣ æt iᴇn ɣ iᴇn w iᴇn wɣ iᴇn ɣ

Qieyun iem iem iep ɣ iep iem iep iem iem ṷiem ṷiep im ɣ im ip ɣ ip ɑn ɑt w ɑn w ɑt ɣ an wɣ an ɣ at wɣ at ɣ an wɣ an ɣ at wɣ at ien ɣ ien w ien wɣ ien ɣ

Input

C1 iəm əm iəp iəp əm əp iəm iəp əm əp im, ʌm (莊) ɨm ip ɨp an al oan, an (幫) oal, al (幫) an oan al oal an oan al oal iən ən, iən iən uən

C2

C3

刷 [soa]

蒜[suən]

C4

品 [pʰum] ɨp

K1

臻

諄 術 殷 迄 文 物

質(櫛)

痕 魂 沒 真(臻)

屑

先

月

元

薛

3

3

合

開

合

3/3A

1 1

4

4

3

3/3A 3B 3/3A 3B 3

3B 3/3A 3B 3

開 合 合 開

合 開 合 開 合

開 合 開

合

開

iɐt ien w ien iet w iet

in it ɣ it w in w it ɨn ɨt iun iut ɣ

ən uon uot in in it ɣ it w in w it ɨn ɨt ɨun ɨut ɣ

ən uon uot in

w

iɤt en w en et w et w

iet iet w iet wɣ iet iɐn w iɐn iɐt ɣ

Input

iᴇt iᴇt w iᴇt wɣ iᴇt iɤn w iɤn iɤt ɣ

Qieyun

əl

巾 [kən]

C1

ɨn, in (幫) il ɨl, il (幫) iun iul ɨn ɨl un ul

ʌn, ɨn on, un ol, ʌl (幫), ul in

uəl, əl, al iən iən iəl iəl

iəl əl iəl iəl ən uən, ən, an əl, al

C2

C3

C4

ɨn ɨl (Continued)

膝 [sɨl]

K1

梗

耕

陌

開 合 開 合 開 合

3

合

蒸 職 蒸 職 庚

1 3

開 合 開

德

2

2

2

1

開 合

登

2

3

3

1

1

曾

合 開 合 開

開 合 開 合 開

江 覺

藥

陽

鐸

唐

江

宕

Table 7.2  (Continued)

iaŋ iak w iak ɣ aŋ ɣ ak

ək ək ɨŋ ɨk w ɨŋ w ɨk ɣ ɐŋ wɣ ɐŋ ɣ ɐk wɣ ɐk ɣ ɐŋ wɣ ɐŋ w

ək ək ɨŋ ɨk w ɨŋ w ɨk ɣ æŋ wɣ æŋ ɣ æk wɣ æk ɣ ɛŋ wɣ ɛŋ w

əŋ əŋ

w

əŋ əŋ

w

w

iaŋ iak w iak ɣ ɔŋ ɣ ɔk w

ɑŋ ɑŋ ɑk w ɑk iaŋ w

Input

ɑŋ ɑŋ ɑk w ɑk iaŋ w

Qieyun

C2

ʌk, ɨk ok, uk ɨŋ ɨk ɨŋ iək ʌiŋ, iəŋ oiŋ ʌik oik ʌiŋ, iəŋ oiŋ

ʌŋ, ɨŋ oŋ, uŋ

aŋ oaŋ ak oak iaŋ, aŋ (幫莊 見) oaŋ iak, ak (見) oak aŋ 鷽學 [hʌk] ak

C1

C3

樁[tsoaŋ]

C4

iŋ ik

K1

通

錫 東 屋 冬 沃 東 屋 鍾 燭

青

昔

陌 清

庚

麥

開 合 開 合 開 開 合 開 合 開 合 開 合 合 合 合 合 合 合 合

1 1 3 3 3 3

4 1

3 3 3 3 3 4

3

2

ɐk ɐk ɣ ieŋ wɣ ieŋ ɣ iek ieŋ w ieŋ iek w iek ieŋ w ieŋ iek oŋ ok oŋ ok iuŋ iuk iṷo̝ŋ iṷo̝k

ɛk ɛk ɣ iæŋ wɣ iæŋ ɣ iæk iᴇŋ w iᴇŋ iᴇk w iᴇk eŋ w eŋ ek uŋ uk oŋ ok iuŋ iuk iṷo̝ŋ iṷo̝k wɣ

wɣ

ɣ

Input

ɣ

Qieyun

C1

C3

ʌik, iək oik iəŋ iəŋ iək ɨk iəŋ iəŋ iək iək iəŋ iəŋ iək ɨk, ik oŋ ok oŋ ok iuŋ, uŋ iuk, uk ioŋ, oŋ, iuŋ iok, ok, uk

C2

C4

K1

218  Conclusion 筮 [siə], 妻 [tsʰiə] and 壻西棲犀 [siə] is dropped due to the dissimilation between the on-glide -i and the off-glide -i. Different from previous studies, this study attributes less readings of SK to nonmainstream layers. First, when SK has two or more correspondences for one MC form, previous studies tended to explain these different representations as different layers which are borrowed from Chinese at different stages. The assumption under this analysis is that the same source form in Chinese must be adapted as the same form in Korean. However, as stated in Chapter 2, the Mandarin loanwords in Modern Korean provide persuasive evidence that the same source form in Chinese can be adapted as different forms in Korean. Therefore, some of the different representations can be explained as a result of misperception or different adaptation strategies. Second, on some occasions, the same rhyme in the source language is represented differently in SK following the place of articulation of initials. For those different representations that are in a complementary distribution, we tend to treat them as reflections of the same layer. It seems highly unlikely that the borrowing of Chinese pronunciations was based on a particular criterion on initials. Third, we think some representations in SK are a result of sound change in Korean. The readings of SK are not affected by the sound change at the same rate. The readings in EMSK include variants in different stages of sound change, and thus some of the irregular readings can be explained by lexical diffusion.

7.3  Process of adaptation of Sino-Korean 7.3.1  Adaptation of initials In this section, we will summarize how the initial and final system of the source language was adapted into Korean and what adjustments were made in the process of adaptation based on the analysis in Chapter 3, Chapter 4 and Chapter 5 of this book. The adaptation of the initials of the source language is given in Table 7.3. One intriguing question about the initial system in EMSK is aspiration assignment, since it is inconsistent with MC. None of the following factors, including rhyme, Table 7.3  Korean adaptation of Chinese initials

Group 幫 (Bilabials)

Group 見 (Velars) Group 影 (Gutturals)

Initial

Input

Adapted form

EMSK

幫(非) 滂(敷) 並(奉) 明(微) 見 溪 群 疑 影 曉 匣 云

p pʰ b m k kʰ g ŋ ʔ h ɦ ø

p p p m k k k ø ø h h ø

p, pʰ p, pʰ p, pʰ m k k k ø ø h h ø

Conclusion 219

Group 端 (Alveolar stops, lateral and nasal) Group 精 (Alveolar fricatives and affricates) Group 知 (Retroflex stops and nasals) Group 莊 (Retroflex fricatives and affricates) Group 章 (Palatal fricatives and affricates)

Initial

Input

Adapted form

EMSK

端 透 定 泥 來 精 清 從 心 邪 知 徹 澄 娘 莊 初 崇 生 章 昌 禪 書 船 日 以

t tʰ d n l ʦ ʦʰ ʣ s z ʈ ʈʰ ɖ ɳ tʂ tʂʰ dʐ ʂ tɕ tɕʰ ʑ ɕ ʑ ȵʑ ø

t t t n l ʦ ʦ ʦ s s t t t n ʦ ʦ ts s ʦ ʦ s s s j ø

t, tʰ tʰ, t t, tʰ n l ʦ, ʦʰ ʦʰ, ʦ ʦ, ʦʰ s s t, tʰ tʰ, t t, tʰ n ʦ, ʦʰ ʦʰ ts, ʦʰ s ʦ, ʦʰ ʦʰ s s s z/j ø

roundness, division or tone in Chinese, is found to be the criterion of aspiration assignment. It is thus necessary to assume that the aspirated initials have not occurred in Korean at the time of borrowing. Later, as Korean started to make the distinction between aspirated and unaspirated sounds, language experts worked to represent the aspiration feature of Chinese more accurately. However, the analogy based on phonetic components often lead them to erroneous readings. In addition, for the convenience of memorization of Chinese character readings, they tended to assign particular syllables with only aspirated readings and others only unaspirated readings. The ri 日 initial was generally represented as [z] in HMJH and HG and was represented as [ø] in YH, CJM, SU, JU and OP. The [z] representation indicates a possibility that the input form of the ri 日 initial was [ʑ], but this also raises the question why had Old Korean adapted the chan 禪 initial and the chuan 船 initial as [s] if it was able to adapt the [ʑ] of the ri 日 initial as [z]. As Lee (1997: 133) suggests, the ri 日 initial ([ȵʑ] in the source language) was probably adapted in Korean as [j] first and later was represented as [z] artificially in HMJH and HG. 7.3.2  Adaptation of finals Table 7.4 is a reconstructed rhyme system of the source language for SK. The principles of reconstruction can be summarized as follows. First, if two rhymes have the same initial and tone but are represented differently in SK, we assume that these two rhymes in the source language are different. For example, the you

u

o o̝ ɤ ɑ ɐ a

e

ə ɨ i

u

o

e

ə

i

a

Vowel

Phoneme

1 3 3 1 2/3 2 3 3/3A 3B 4 1 3 3A 3B

1 3

Div.

au 肴

i 支脂 ɣ i 支脂

ieu 宵 ɣ ieu 宵 ieu 蕭

ɣ

a麻

ɣ

ɑu 豪

ɨu 尤 iu 幽

ɑ歌

o模

u侯

no coda

ia 麻

iu̯ o̝ 虞 iɤ 魚

iei 祭廢 ɣ iei 祭廢 iei 齊 əi 咍 ɨi 之微

ɑi 泰 ɐi皆佳 ɣ ai 夬 ɣ

Table 7.4  Reconstructed rhyme system of the source language of SK

iek 錫 ək 德 ɨk 職

iek 陌

ɑk 鐸 ɐk 陌麥 ɣ ak 覺 ɣ

ok 沃屋1 iu̯ o̝ k 燭

iuk 屋3

ɨt 迄 it 質 ɣ it 質

iet 薛 ɣ iet 薛 iet 屑

ɑt 曷 iɐt 月 ɣ at 黠鎋

iut 物

-k, -t, -p coda

ip 緝 ip 緝 ɣ

iep 葉業 ɣ iep 葉 iep 帖 əp 合

ap 洽狎 ɣ

ɑp 盍

ieŋ 青 əŋ 登 ɨŋ 蒸

ɑŋ 唐 ɐŋ 庚耕 ɣ aŋ 江 iaŋ 陽 ieŋ 庚清 ɣ

oŋ 冬東1 iu̯ o̝ ŋ 鍾

iuŋ 東

ien 仙 ɣ ien 仙 ien 先 ən 痕 ɨn 殷 in 真 ɣ in 真

im 侵 im 侵 ɣ

iem 鹽嚴 ɣ iem 鹽 iem 添 əm 覃

ɑn 寒 ɑm 談 iɐn 元 ɣ an 山刪 ɣam 咸銜

on 魂

ɨun 文

-ŋ, -n, -m coda

Conclusion 221 尤 rhyme character jiu 九 and the you 幽 rhyme character jiu 糾 share the same initial (the jian 見 initial) and tone (the shang 上 tone). In SK, the former one is represented as [ku], while the latter one as [kiu]. To distinguish these two rhymes, we reconstructed the you 尤 rhyme as [ɨu] and the you 幽 rhyme as [iu]. Second, the reconstructed form should provide a reasonable explanation for the various representations in SK. For example, in order to explain the [ən]/[an] representations in the yuan 元 rhyme and the [əl]/[al] representations in the yue 月 rhyme, we reconstruct the nuclear vowel of these two rhymes in the source language as [ɐ]. Third, the reconstructed form in the source language should be consistent with the general trend of rhyme development in Chinese. For example, Huilin’s fanqie in Yiqie jing yinyi shows that some double rhymes have already merged into one type in Chang’an dialect around the ninth century, so if two double rhymes have exactly the same representation in SK, we will reconstruct the same phonetic value for them if no counter evidence can be found. It should be noted that this could be too hasty, since it is probable that there may be a distinction between the two rhymes in the source language, while not recognizable from the adapted form in SK. Nevertheless, before we can find a clue about the difference among these rhymes, this reconstruction can give us a general idea about the rhyme system of the source language. There are six vocalic phonemes in the source language: [u, o, a, e, ə, i], from which 11 allophones branch out. We propose that the second division has [ɣ] medial in this language since this medial [ɣ] occurs as [i] in SK in some cases, such as the [io] reading of the yao 肴 rhyme. This language makes distinction in chongniu. Div. 3/3A has -i medial and div. 3B has -ɣi medial. Fourth division has merged into div. 3 and includes an -i medial. The adaptation of the vowels of the source language is given in Table 7.5. Table 7.5  Korean adaptation of Chinese vowels Phoneme

Input

Adapted form

EMSK

u o

u o o̝ ɤ ɑ ɐ a e ɨ ə i

u o u, o ə, o (莊) a ə, ʌ, a a e ɨ ɨ, ə i

u o u, o ə, o (莊) a ə, ʌ, a a ə ɨ ɨ, ə, ʌ, i i

a e ə i

The [u] vowel in the source language was adapted as [u], as in 侯 [u] > [u], 幽 [iu] > [iu], 尤 [ɨu] > [iu]/[u], 文 [ɨun] > [un], 物 [ɨut] > [ul]. The [o] vowel in

222  Conclusion the source language has three allophones, namely [o], [o̝ ] and [ɤ]. The [o] sound was generally adapted as [o]: 模 [ṷo] > [o], 冬東 (Div. 1) [oŋ] > [oŋ], 沃屋 (Div. 1) [ok] > [ok], 魂 [uon] > [on], 沒 [uot] > [ol]/[ul]/[ʌl](labials). The [o̝] sound is articulated with the tongue raised higher than [o] and was adapted as either [u] or [o]: 虞 [iṷo̝] > [iu]/[u], 鍾 [iuᴖ o̝ ŋ] > [ioŋ]/[oŋ]/[uŋ], 燭 [iuᴖ o̝ k] > [iok]/[ok]/[uk]. The [ɤ] sound was adapted as [ə], but as [o] after retroflexes, as in 魚 [iɤ]>[iə]/ [ə]/[o](retroflexes). The [a] vowel in the source language has three allophones, namely [ɑ], [ɐ] and [a]. The [ɑ] sound was adapted as [a] except in the hao 豪 rhyme: 歌 [ɑ] > [a], 泰 [ɑi] > [ai], 鐸 [ɑk] > [ak], 曷 [ɑt] > [al], 盍 [ɑp] > [ap], 唐 [ɑŋ] > [aŋ], 寒 [ɑn] > [an], 談 [ɑm] > [am]. Due to the phonotactic constraint in Korean which does not allow [ɑu], the hao 豪 rhyme [ɑu] was adapted as monophthong [o]. The [ɐ] sound was adapted as [ə] or [a] for third division rhymes: 月 [iɐt] > [əl]/[al], 元 [iɐn] > [ən]. For second division rhymes 皆佳庚 (2) 耕陌麥, there is a [ɣ] medial. Generally speaking, the [ɣ] medial is not clearly perceived and is dropped in the adapted form, and the [ɐ] sound is adapted as [ʌ]. For velar/guttural initials, the [ɣ] medial in some characters is more salient and was adapted as [i], and the [ɐ] sound was more front and high and thus was adapted as [ə]. The [a] sound was adapted as [a], as in 麻 [ɣa] > [a], 肴 [ɣau] > [io]/[o], 夬 [ɣai] > [ai], 覺 [ɣak] > [ak], 黠鎋 [ɣat] > [al], 洽狎 [ɣap] > [ap], 江 [ɣaŋ] > [aŋ], 山刪 [ɣan] > [an], 咸銜 [ɣam] > [am], 麻 [ia] > [ia], 陽 [iaŋ] > [iaŋ]. The [e] vowel in the source language was generally adapted as [e] and later became [ə]. Based on the evidence from Mongolian loanwords and the record in the book Jilin leishi, Lee and Ramsey (2011: 95) proposed that the vowel shift /e/>/ə/ occurred in Korean between the thirteenth and fifteenth centuries. 宵蕭 [ieu] is adapted as [io] due to the phonotactic constraint in Korean which does not Table 7.6  Korean adaptation of Chinese [e]

iei 祭廢齊 ien 仙先 iet 薛屑 iem 鹽嚴添 iep 葉業帖 ieŋ 庚 (3) 清青 iek 錫陌(3)

Adapted form

Sound change

EMSK

iei ien iel iem iep ieŋ iek

iəi iən iəl iəm iəp iəŋ iək

iəi iən iəl iəm iəp iəŋ iək

allow [eu]. Table 7.6 summarizes the process through which the adapted forms became the forms in EMSK. The [ə] vowel in the source language has two allophones: [ə] and [ɨ]. The [ə] vowel was adapted as either [ə] (and later generally changed to [ʌ]) or [ɨ]. The [ɨ] vowel was adapted as [ɨ] and later changed to [i] in some instances.

Conclusion 223 Table 7.7  Korean adaptation of Chinese [ə]

ək 德 əŋ 登 ən 痕 əi 咍 əm 覃 əp 合 ɨi 之微 ɨn 殷 ɨt 迄 ɨŋ 蒸 ɨk 職

Adapted form

Sound change

EMSK

ək ɨk ɨŋ ən ɨn əi əm əp ɨi ɨn ɨl ɨŋ ɨk

ʌk

ʌk, ək, ɨk

ʌn

ɨŋ ʌn, ɨn

ʌi > ai ʌm > am ʌp > ap i iŋ ik

ʌi, ai ʌm, am ap ɨi, i ɨn ɨl ɨŋ, iŋ ɨk, ik

The [i] vowel in the source language was adapted as [i]: 支脂 [i] > [i], 質 [it] > [il], 緝 [ip] > [ip], 真 [in] > [in], 侵 [im] > [im]. 7.3.3  Factors involved in adaptation 7.3.3.1  Phonotactic constraints in Korean In the process of adaptation, those sounds or phonological structures that are not available in the target language will be adjusted to fit the phonotactic constraints of the adapting language. For example, the yi 疑 [ŋ] initial, the ying 影 [ʔ] initial and the ri 日 [ȵʑ] initial are represented as a zero-initial. The zhuang 莊 series (retroflex affricates and fricatives), the zhang 章 series (palatal affricates and fricatives), and the jing 精 series (alveolar affricates and fricatives) are represented as one type, namely alveolar affricates and fricatives. In EMSK, the huan 桓 rhyme [wɑn] is represented as [an] for alveolar initials. This can also be found in Eom’s (1996, 2002) transcription of Mandarin loanwords, in which representations that are rarely used in native Korean phonology, such as [twɑn] and [thwɑn], are avoided. The ge 戈 [ʷɑ] rhyme (hekou div. 1) is represented as [a] or [oa] generally according to the articulation place of initials. For alveolar stops, laterals and nasals, div. 1 戈 [ʷɑ] is represented as [a] due to the phonotactic constraint in Korean, which does not allow syllable types like [toa], [tʰoa], [noa] and [loa]. The syllable structure in Middle Chinese, as shown in the Qieyun, can be expressed as “IMGVE,” in which “I” refers to Initial, “M” Medial, “G” on-glide, “V” nuclear vowel and “E” Ending (Pan and Zhang 2013). The syllable structure in Middle Korean is “IMVF,” in which “I” refers to Initial, “M” Medial, “V” nuclear vowel and “F” Final vowel (Kōno 1979). The number of elements

224  Conclusion between the initial and the nuclear vowel could be two in MC, while only one element is allowed in this position in ancient Korean. In other words, the glides in Middle Chinese were invariably simplified through the process of adaptation. For example, the hekou ji 祭 rhyme [ʷiei] is generally represented as [iəi] in EMSK. 7.3.3.2  Role of perception The same form in Chinese could be adapted as different forms in Korean. One of the reasons is the influence of perception. The phonetic value of the same rhyme could vary when it is followed by different initials, and this subtle difference could cause the same phoneme to be adapted differently in the target language. The yu 魚 rhyme [iɤ] is generally represented as [ə] when the initial is velar or guttural, [o] when the initial is a retroflex fricative or affricate, and [iə] elsewhere. This [o] is explained as a reflection of the [ï̱ o] in Old Chinese in Pak (1971) and Lee (1997) and as a borrowing of Old Mandarin in Arisaka (1936) and Kōno (1979). However, the representation of the yu 魚 rhyme [iɤ] varies following the place of articulation. It is unlikely that the reading of Chinese characters is borrowed based on a particular phonological criterion. This study argues that the main vowel of the yu 魚 rhyme become more back and rounded under the influence of the zhuang initials, which are retroflexes, and thus borrowers tend to use [o], which has [+back] and [+round] features, to transcribe this sound. The borrowers’ perception of sounds is also affected by the tone. Kōno (1979: 463–464) noticed that in hekou yuan 元 rhyme, for labial initials ping tone is usually represented as [ən], shang tone and qu tone as [an], and ru tone as either [əl] or [al]. This tendency is generally also obeyed in Sino-Vietnamese (Ito 2007: 186). The [i] glide of the hekou yuan 元 rhyme [wiɐn] can be clearly perceived when it is a ping tone, but not so when it is a shang tone or a qu tone. This difference causes the input being adapted differently. 7.3.3.3  Different adaptation strategies The discussion on Korean transcription of Mandarin loanwords in Chapter 2 gives us the following hints. First, what constitutes the most suitable or faithful way to transcribe loanwords could be different depending on the perspective of the borrower. For instance, the -o in -ong is represented as [uŋ] in government transcription, while it is represented as [oŋ] in Choe-Kim (1985) and Eom (2002). Second, the same translator may propose different ranking in different words. Loanword borrowing, even when observed by the same borrower, could present different rankings of constraints, and consequently different results. For example, the -o in -ao and -o in -ong in Mandarin are both [u]. However, in the government transcription the former one is represented as [a.o], while the latter one as [uŋ]. Although there is no historical record about the process of transmitting Chinese readings into ancient Korean, this task is not likely to be completed by one single person. There could be many inconsistencies among different borrowers’ patterns

Conclusion 225 of borrowing. Due to different adaptation strategies, the same input form in Chinese was occasionally adapted as two or three different forms in SK irrespective of phonological environment. The zhu 燭 rhyme is generally represented as [iok] or [ok] in EMSK, but the reading of 局 and 旭 is [kuk] and [uk], respectively. The phonetic reconstruction for the zhu 燭 rhyme in the source language is [jṷo̝k]. Since the [o̝ ] in [jṷo̝k] has a higher tongue position than [o̝ ], it may be perceived as [u] by borrowers, and thus the rhyme of 局 and 旭 was adapted as [uk].

7.4  Irregular readings in Sino-Korean The primary reason for irregular reading is the analogy of phonetic radicals, such as 吉 (質 rhyme) > 桔䓀 [kil], 危 (支 rhyme) > 桅 [ui], 者 (麻 rhyme) > 煮翥 [tsia], 粹 (脂 rhyme) > 焠 [siu], 螘 (支 rhyme) > 磑 [ɨi], 叉 (麻 rhyme) > 釵 [tsʰa], 氐底低 (齊 rhyme) > 䑛 [tiəi], 台 (咍 rhyme) > 笞 [tʰʌi], 宰 (咍 rhyme) > 滓 [tsʌi], 丹 (寒 rhyme) > 坍 [tan], 鷸繘 (術 rhyme) > 譎 [hiul] and so forth. It is noteworthy that the result of analogy may not necessarily cause the target character to be read exactly the same as the source of analogy. In some cases, only part of the reading is affected by analogy. This is called contamination in Ito (2007). For example, the initial of 槲斛 [kok] is probably contaminated from the initial of 角 [kak], and the initial of 莖 [kʌiŋ] results from contamination of the initial of 經 (見 initial) [kiəŋ] or 輕 (溪 initial) [kiəŋ]. Another possible cause of irregular readings is the influence of rhyme books. We can safe to claim that not all the readings of Chinese characters in SK are borrowed from Chinese directly. This is especially true in the case of rarely used characters. Therefore, by surveying the rhyme books used in ancient Korea, we may answer the origin of some irregular readings in EMSK. For example, the reading of 隘 [ʌik] is consistent with the yige 乙革 fanqie spelling in the rhyme book HWZY.

7.5  Future work This book outlines a general framework of the origin, layers and adaptation of SK. Some conclusions reached in this book are controversial and thus await further research. After the adaptation, subsequent sound change in Korean must have changed the readings in SK to a certain extent. In this study, we have proposed the possible stages through which the adapted form changed to EMSK generally based on alphabetic materials after the fifteenth century through the application of internal reconstruction. To ensure its validity, more attention will be paid to the investigation of those documents before the invention of the Korean alphabet in our future work. In this book, we explained the irregular readings in many cases as a result of analogy. It is necessary to point out that the analogy analysis provides a possible explanation, but it could be no more than a speculation of modern scholars if we do not have support from other evidence. Analogy might well explain most cases

226  Conclusion of irregular readings, but we have to be alert of the possibility of overusing this kind of explanation because the majority of Chinese characters are phonograms and it is relatively easy to set up a connection between two characters which share the same phonetic radical. In addition, we explained the readings of some characters as a reflection of Old Chinese, such as 鷽學 [hʌk], 巾 [kən] based on the similarity between those forms and the reconstructed forms of Old Chinese. However, there is always no decisive evidence to claim these readings must be borrowed from Old Chinese. These readings may be results of some other factors, which we have noticed, or some factors which we are not aware of yet. To have more certainty, we will do more research to test these explanations in the future.

References

Arisaka, Hideyo. 1936. “Kanji no Chōsen’on ni tsuite [On the Korean reading of Chinese characters].” Hōgen, 6, 4: 16–25. ———. 1957. Kokugo on’inshi no kenkyū [A Study on the History of Japanese Phonology]. Tokyo: Sanseidō. Baxter, William. 1992. A Handbook of Old Chinese Phonology. New York: Mouton de Gruyter. Chao, Yuen Ren. 1930. A system of tone letters. Le Maître Phonétique, 45: 24–27. ———. 1948. The voiced velar fricative as an initial in Mandarin. Le Maître Phonétique, 89: 2–3. ———. 1968. A Grammar of Spoken Chinese. Berkeley: University of California Press. Cheng, Chin-chuan. 1973. A Synchronic Phonology of Mandarin Chinese. The Hague: Mouton. Choe, Hyewon. 2002. Pyojuneo baleum siltae josa [A Survey on the Pronunciation of Standard Korean]. Seoul: Guglib gugeowon. Davidson, Lisa and Rolf Noyer. 1997. Loan phonology in Huave: nativization and the ranking of faithfulness constraints. Proceedings of the West Coast Conference on Formal Linguistics, 15: 65–79. Dong Tonghe. 1996. Hanyu yinyunxue [Chinese Historical Phonology]. Taibei: Wenshizhe chubanshe. Duanmu, San. 1990. A Formal Study of Syllable, Tone, Stress and Domain in Chinese Languages. Ph.D. dissertation. Cambridge: Massachusetts Institute of Technology. ———. 2000. The Phonology of Standard Chinese. New York: Oxford University Press. Dupoux, Emmanuel, Kazuhiko Kakehi, Yuki Hirose, Christophe Pallier, and Jacques Mehler. 1999. Epenthetic vowels in Japanese: a perceptual illusion? Journal of Experimental Psychology: Human Perception and Performance, 25: 1568–1578. Eom, Ik-sang. 1991. A Comparative Phonology of Chinese and Sino-Paekche Korean. PhD dissertation. Indiana University. ———. 1996. “Junggugeo hangeul pyogibeobui munjejeomgwa gaeseon bang’an [The problem in the Koreanization system of Mandarin Chinese and a revised proposal].” Junggug eoneo yeongu [Journal of Chinese Linguistics in Korea], 4: 39–84. ———. 1999. Different layers of borrowing: Sino-Korean characters with multiple readings. Journal of Chinese Linguistics, 27, 2: 27–44. ———. 2001. “Hanjaeum -l unmi junggug bangeum giwonseolui munje [Problems of tracing the origin of Sino-Korean coda -l from Chinese dialect].” Jungeo Jungmunhag [Korea Journal of Chinese Language and Literature], 29: 21–35.

228  References ———. 2002. Junggug eoneohag hangugsigeulo hagi [Chinese Linguistics from a Korean Perspective]. Seoul: Hangug munhwasa. ———. 2003. The mechanism of Sino-Korean palatalization. Journal of Chinese Linguistics, Monograph Series No. 20, The Lexical Diffusion of Sound Change in Korean and Sino-Korean: 175–215. ———. 2008. Hangug hanjaeum junggugsigeulo bogi [Sino-Korean Phonology from a Chinese Linguistic Perspective]. Seoul: Hangug munhwasa. ———. 2015. Hangug hanjaeum junggugsigeulo dasi bogi [Sino-Korean Phonology Revisited from a Chinese Linguistic Perspective]. Seoul: Hangug munhwasa. Fu, Maoji. 1956. “Beijing hua de yinwei he pinyin zimu [Phonemes and Pinyin symbols in the Beijing speech].” Zhongguo yuwen [Studies of the Chinese Language], 5: 3–12. Hu, Mingyang. 1963. “Laoqida yanjie he piaotongshi yanjie zhong suojian de hanyu chaoxianyu duiyin [The Korean transcription of Chinese in Laoqida yanjie and Piaotongshi yanjie].” Zhongguo yuwen [Studies of the Chinese Language], 3: 185–192. Huang, Cuibo. 1931. Huilin yiqie jing yinyi fanqie kao [A Study on the fanqie in Huilin’s Yiqie jing yinyi]. Institute of History and Philology Monograph Series No. 6. Shanghai: Institute of History and Philology, Academia Sinica. International Phonetic Association, eds. 1999. Handbook of the International Phonetic Association: A Guide to the Use of the International Phonetic Alphabet. Cambridge: Cambridge University Press. Ito, Chiyuki. 2007. Chōsen Kanjion kenkyū [Sino-Korean Phonology]. Tokyo: Kyūko syoin. Ito, Chiyuki and Michael Kenstowicz. 2009. Mandarin loanwords in Yanbian Korean II: tones. Language Research, 45, 1: 85–109. Itô, Junko and Armin Mester. 1995. The core-periphery structure of the lexicon and constraints on reranking. In Jill N. Beckman, Laura Walsh Dickey, and Suzanne Urbanczyk (eds.), Papers in Optimality Theory: University of Massachusetts Occasional Papers in Linguistics, 18: 181–209. Amherst: GLSA. Iverson, Gregory K. and Ahrong Lee. 2006. Perception of contrast in Korean loanword adaptation. Korean Linguistics, 13: 49–87. Jacobs, Haike and Carlos Gussenhoven. 2000. Loan phonology: perception, salience, the lexicon and OT. In Joost Dekkers, Frank van der Leeuw, and Jeroen van de vweijer (eds.), Optimality Theory: Phonology, Syntax and Acquisition, 193–210. Oxford: Oxford University Press. Jeon, Gwangjin. 1999. “Junggugeo jaeumui hangeul pyogibeobe daehan eumseonghakjeok daebi bunseok [Phonetical analysis on Sino-Korean notating systems for initial consonants].” Junggug munhag yeongu [Journal of Chinese Literature], 19: 347–375. Ji, Xianlin. 1985. Datang xiyuji jiaozhu [The Collation and Annotation of the Great Tang Dynasty Record of the Western Regions]. Beijing: Zhonghua shuju. Jin, Youjing. 1998.“Hanyu shi shang -ï yin de chansheng niandai [The appearing time of [ï] in Chinese].” Xuzhou shifan daxue xuebao [Journal of Xuzhou Normal University], 24, 3: 57–60. Kang, Sinhang. 2003. Hanhan eumunsa yeongu [A Study on the History of Sino-Korean Phonology]. Seoul: Taehagsa. ———. 2010. “Hangug hanjaeumui laewongwa myeoch gaji teugseonge daehayeo [On the origin and several features of Sino-Korean].” Proceeding of the 40th National Conference of the Kugyol Association: 78–99. Karlgren, Bernhard. 1922. “The reconstruction of Ancient Chinese.” T’oung Pao, 21: 1–42. ———. 1926 [1940]. Etudes sur la phonologie chinoise. Archaic d’etudes orientales, vol. 15. Chinese translation by Chao, Yuenren, Fangkuei Li, and Changpei Luo. Zhongguo

References 229 yinyunxue yanjiu [A Study on Chinese Historical Phonology]. Shanghai: Shangwu yinshuguan. ———. 1954. Compendium of phonetics in Ancient and Archaic Chinese. Bulletin of the Museum of Far Eastern Antiquities, 26: 211–367. Kenstowicz, Michael. 2003. The role of perception in loanword phonology. Studies in Africa Linguistics, 32: 95–112. Kim, Mi-ryoung and San Duanmu. 2004. Tense and lax stops in Korean. Journal of East Asian Linguistics, 13: 59–104. Kim, Tae-eun. 2012. A Study of Mandarin Loanwords: Lexical Stratification, Adaptation and Factors. Ph.D. dissertation. Madison: University of Wisconsin. Kim, Yong-ok. 1985. Dongyanghag eotteohge hal geos inga [How to Do Oriental Studies]. Seoul: Mineumsa. Kōno, Rokurō. 1979. Kōno Rokurō chosakushū [The Collected Works of Kōno Rokurō]. Tokyo: Heibonsha. Kwon, Inhan. 2005. Jungse hangug hanja eumhun jipseong [A Collection of the Sound and Meaning of Chinese Characters in Middle Sino-Korean]. Seoul: Jeiaenssi. LaCharité, Darlene and Carole Paradis. 2005. Category preservation and proximity versus phonetic approximation in loanword adaptation. Linguistic Inquiry, 36: 223–258. Ladefoged, Peter and Ian Maddieson. 1996. The Sounds of the World’s Languages. Oxford: Blackwell Publishers. Lee, Donju. 1980. “Hunmong jahoe hanjaeumeseo balgyeondoen junggugeumui yeonghyange daehayeo [On the influence of Chinese in the reading of Hunmong jahoe].” Ilsan gim junyeong seonsaeng hwagap ginyeom nonchong: 117–130. ———. 1990. Hunmong jahoe hanjaeumn yeongu [A Study on the Sino-Korean System in Hunmong jahoe]. Seoul: Hongmungak. Lee, Haewoo. 1994. An Etymophonology Comparison of Chinese Dialects, Sino-Japanese, and Sino-Korean. Ph.D. dissertation. University of Hawaii. Lee, Iksop and S. Robert Ramsey. 2000. The Korean Language. New York: State University of New York Press. Lee, Juhaeng and Sangjun Kim. 2010. Aleumdaun hangugeo [Beautiful Korean Language]. Revised version. Paju: Jigu munhwasa. Lee, Ki-Moon. 1963. Gugeo pyogibeobui yeogsajeog yeongu [Historical Studies on the Korean Writing System]. Seoul: Korean Research Center. ———. 1972. Gugeosa gaeseol [An Introduction to the History of Korean Language]. Revised edition [2004], Seoul: Taehagsa. Lee, Ki-Moon and S. Robert Ramsey. 2011. A History of the Korean Language. Cambridge: Cambridge University Press. Lee, Yundong. 1997. Hangug hanjaeumui ihae [On Sino-Korean Phonology]. Seoul: Hyeongseol chulpansa. Li, Dechun. 1985. “Hanyu shangguyin zai shiliu shiji chaoxian hanziyin zhong de yicun [The remains of Old Chinese in the 16th century Sino-Korean].” Minzu yuwen [Minority Languages of China], 5: 36–39. ———. 1994. Joseoneo hanjaeoeum yeongu [A Study on the Reading of Chinese Characters in Korean]. Seoul: Sekwang haksul jalyosa. Li, Xinkui. 2000. Zhonggu yin [Middle Chinese Phonology]. Beijing: Shangwu yinshuguan. Liao, Rongrong and Feng Shi. 1987. Hanyu putonghua -r shengmu yinzhi de shiyan yanjiu [An experimental study on the phonetic value of -r initial in Mandarin]. In Yuyin conggao [Collected Papers on Phonetics], 177–197. Beijing: Beijing yuyan xueyuan chubanshe.

230  References Lin, Yen-Hwei. 2007. The Sounds of Chinese. Cambridge: Cambridge University Press. ———. 2009. Loanword adaptation and phonological theory. In Yun Xiao (ed.), Proceedings of the 21st North American Conference on Chinese Linguistics, 1–12. Smithfield, Rhode Island: Bryan University. Lu, Zhiwei. 1946. “Shi Zhongyuan yinyun [on Zhongyuan yinyun].” Yanjing xuebao [Yenching Journal of Chinese Studies], 31: 1–36. Luo, Changpei. 1933. Tang Wudai xibei fangyin [The North-Western Dialect in Tang and Five Dynasties]. Institute of History and Philology Monograph Series A, No. 12. Shanghai: Institute of History and Philology, Academia Sinica. Martin, Samuel. 1997. “How did Korean get -l for Middle Chinese words ending in -t?” Journal of East Asian Linguistics, 6: 263–271. Maspero, Henri. 1920 [2005]. “Les dialecte de Tch’ang-ngan sous les T’ang [Chang’an dialect of the Tang period].” Bulletin de l’Ecole Francaise d’Extreme-Orient, 20. Chinese translation by Nie, Hongyin. Tangdai Chang’an fangyan kao [A Study on Chang’an Dialect in the Tang Dynasty]. Beijing: Zhonghua shuju. Miao, Ruiqin. 2005. Loanword Adaptation in Mandarin Chinese: Perceptual, Phonological and Sociolinguistic Factors. Ph.D. dissertation. Stony Brook University. Nam, Gwangu. 1973. Joseon ijo hanjaeum yeongu [A Study on the Sino-Korean Pronunciations in Joseon]. Seoul: Iljogak. Nie, Hongyin. 1984. “Qieyun chongniu san si deng zi de chaoxian duyin [The Korean reading of the third and fourth division double rhyme characters in Qieyun].” Minzu yuwen [Minority Languages of China], 3: 61–66. Ning, Jifu. 1985. Zhongyuan yinyun biaogao [Draft Table of Zhongyuan yinyun]. Changchun: Jilin wenshi chubanshe. Norman, Jerry. 1988. Chinese. Cambridge: Cambridge University Press. Pak, Byeongchae. 1971. Godae gugeoui yeongu [A Study on Ancient Korean]. Seoul: Korea University press. Pak, Jeongbong. 2009. Hanguo hanzi yinxi yu Xiamen fangyan yinxi bijiao yanjiu [A Comparative Study on Sino-Korean Phonology and Xiameness Phonology]. Chengdu: Bashu shushe. Pan, Wuyun. 2000. Hanyu lishi yinyunxue [Chinese Historical Phonology]. Shanghai: Shanghai jiaoyu chubanshe. Pan, Wuyun and Hongming Zhang. 2013. “Hanyu zhongguyin [Middle Chinese phonology].” Yuyan yanjiu [Studies in Language and Linguistics], 33, 2: 1–7. Paradis, Carole. 1996. The inadequacy of filters and faithfulness in loanword adaptation. In Jacques Durand and Bernhard Laks (eds.), Current Trends in Phonology: Models and Methods, 509–534. Salford: University of Salford Publication. Paradis, Carole and Darlene LaCharité. 1997. Preservation and minimality in loanword adaptation. Journal of Linguistics, 33: 379–430. Peperkamp, Sharon. 2005. A psycholinguistic theory of loanword adaptation. Proceedings of the 30th Annual Meeting of the Berkeley Linguistics Society: 341–352. Peperkamp, Sharon and Emmanuel Dupoux. 2003. Reinterpreting loanword adaptation: the role of perception. Proceedings of the 15th International Congress of Phonetic Science: 367–370. Peperkamp, Sharon, Inga Vendelin, and Kimihiro Nakamura. 2008. On the perceptual origin of loanword adaptation: experimental evidence from Japanese. Phonology, 25: 129–164. Pulleyblank, Edwin G. 1983. Vowelless Chinese? an application of the Tree-tired theory of syllable structure. Proceedings of the Sixteenth International Conference on SinoTibetan Languages and Linguistics. Seattle: University of Washington. 568–619.

References 231 ———. 1984. Middle Chinese: A Study in Historical Phonology. Vancouver: University of British Columbia Press. Shao, Rongfen. 1981. “Jinshu yinyi fanqie de yuyin xitong [The sound system in the fanqie spellings of Jinshu yinyi].” Yuyan yanjiu [Studies in Language and Linguistics], 1: 103–124. Shen, Dongyue. 2004. “Hanhan duiyin zhong weihe *-t＞-l [The reason of *-t>-1 in SinoKorean].” Minzu yuwen [Minority Languages of China], 4: 31–36. Shi, Xiangdong. 1983. “Xuanzang yizhu zhong de fanhan duiyin yu Tang chu zhongyuan fangyin [The Sanskrit-Chinese transcription in Xuanzang’s translations and the phonological system of Zhongyuan dialect in early Tang].” Yuyan yanjiu [Studies in Language and Linguistics], 1: 27–48. Shin, Jiyoung, Jieun Kiaer, and Jaeeun Cha. 2012. The Sounds of Korean. Cambridge: Cambridge University Press. Shin, Woosun. 2015. Hanguo hanziyin lishi cengci yanjiu [A Study on the Diachronic Strata of Sino-Korean Pronunciations]. Ph.D. dissertation. National Taiwan University. Silverman, Daniel. 1992. Multiple scansions in loanword phonology: evidence from Cantonese. Phonology, 9: 289–328. Smith, Jennifer L. 2006. Loan phonology is not all perception: evidence from Japanese loan doublets. Japanese and Korean Linguistics, 14: 63–74. Sohn, Ho-Min. 1999. The Korean Language. Cambridge: Cambridge University Press. Synn, Asa. 2006. Han han yin yanjiu [A Study on Sino-Korean Phonology]. Ph.D. dissertation. Peking University. Takagi, Naoyuki and Virginia Mann. 1994. A perceptual basis for the systematic phonological correspondences between Japanese loan words and their English source words. Journal of Phonetics, 22: 343–356. Vendelin, Inga and Sharon Peperkamp. 2004. Evidence for phonetic adaptation of loanwords: an experimental study. Actes des journées d’Etudes Linguistiques: 129–131. ———. 2006. The influence of orthography on loanword adaptation. Lingua, 116: 996–1007. Wang, Hongzhi. 2012. “Zaoqi zhisi yun de yinzhi jiqi yanjin [Early sound values and historical evolution of the 「zhi-si」 rime category in Chinese].” Journal of Chinese Linguistics, 40, 2: 420–441. Wang, Li. 1948. “Hanyueyu yanjiu [A study of Sino-Vietnamese].” Lingnan xuebao [Lingnan Journal of Chinese Studies], 9, 1: 1–96. ———. 1982a. “Xuanying yiqie jing yinyi fanqie kao [A study of the fanqie spellings in Xuanying’s Yiqie jing yinyi].” Yuyan yanjiu [Studies in Language and Linguistics], 1: 1–5. ———. 1982b. “Jingdian shiwen fanqie kao [A study of the fanqie spellings in Jingdian shiwen].” Longchong bingdiao zhai wenji [Collection of Essays from Longchong Bingdiao Studio], 3: 135–211. Beijing: Zhonghua shuju. ———. 1982c. “Zhu Xi fanqie kao [A study of Zhu Xi’s fanqie spellings].” Longchong bingdiao zhai wenji [Collection of Essays from Longchong Bingdiao Studio], 3: 257–338. Beijing: Zhonghua shuju. ———. 1983. “Zailun rimu de yinzhi jianlun putonghua shengmu biao [A reexamination of the phonetic value of ri initial: a discussion on the initial inventory in Mandarin].” Zhongguo yuwen [Studies of the Chinese Language], 1: 20–23. ———. 1985. Hanyu yuyin shi [Chinese Historical Phonology]. Beijing: Zhongguo shehui kexue chubanshe. Yang, Naisi. 1981. Zhongyuan yinyun yinxi [The Phonology of Zhongyuan yinyun]. Beijing: Zhongguo shehui kexue chubanshe.

232  References Yip, Moira. 1993. Cantonese loanword phonology and optimality theory. Journal of East Asian Linguistics, 9: 261–291. ———. 2002. Necessary but not sufficient: perceptual loanword influence in loanword phonology. The Journal of the Phonetic Society of Japan, Special Issue on Aspects of Loanword Phonology, 4: 4–21. ———. 2006. The symbiosis between perception and grammar in loanword phonology. Lingua, 116: 950–975. Yu, Changgyun. 1980. Hangug godae hanjaeum ui yeongu I [A Study on Ancient SinoKorean Phonology I]. Daegu: Gyemyeong University Press. ———.1991. Samgug sidae ui hanjaeum [The Pronunciation of Sino-Korean in the Three Kingdoms Period]. Seoul: Mineumsa. Yu, Hui. 2014. Hanyu jieci yinxixue – yingyuan jieci de yuyin he yinxi fenxi [Chinese Loanword Phonology: A Phonetic and Phonological Analysis of English Loanword Adaptation in Chinese]. Tianjin: Nankai University Press. Zhang, Weijia. 2008. “Chaoxian hanziyin de cengci he laili [The layer and source of SinoKorean phonology].” Minzu yuwen [Minority Languages of China], 5: 3–16. Zhang, Yi. 1994. Wang wu tianzhuguo zhuan jianzhu [Annotation on the Memoir of the Pilgrimage to the Five Kingdoms of India]. Beijing: Zhonghua shuju. Zhou, Fagao. 1948. “Xuanying fanqie kao [A study of Xuanying’s fanqie spellings].” Bulletin of the Institute of History and Philology, Academia Sinica, 20: 359–444. ———. 1989. “Sui Tang Wudai Song chu chongniu fanqie yanjiu [A study of the fanqie spellings of chongniu rhymes in Sui, Tang, Five Dynasties and early Song].” Proceedings of the Second International Conference on Sinology, Academia Sinica: 85–109. Zhou, Zumo. 1966. “Song dai Bianluo yuyin kao [Research on Bianluo sounds in the Song Dynasty].” Wenxue ji [Collections of Seeking Knowledge]. Beijing: Zhonghua shuju. Zhu, Xiaonong. 1982. “Guanyu putonghua rimu de yinzhi [On the phonetic value of Mandarin ri initial].” Zhongguo yuwen tongxun [Newsletter of Chinese Language], 3: 19–22.

Index

adaptation strategy 16, 24 analogy 12, 15 – 16, 62 – 63, 78 – 79, 88, 94, 103, 105 – 106, 109 – 110, 112 – 114, 116 – 117, 119 – 120, 122 – 124, 127 – 129, 131 – 133, 136, 138, 140, 142, 144 – 145, 148, 150 – 155, 157 – 158, 160, 162 – 164, 168 – 169, 171, 174, 176, 178 – 179, 191, 220, 225 – 226 apical 26 – 27 assimilation 31, 79, 103, 113

idu 吏讀 1, 3 inner source layer 16 internal reconstruction 226 irregular reading 12, 219, 225 – 226

Buddhist 2, 4, 13, 19 – 22, 205 – 207

lenition 63, 78, 183 lexical diffusion 219 lexical tone 30 Loanword Phonology 1, 15 – 16, 22, 24, 49 – 50, 56

chitou 齒頭 13 chongniu 重紐 1, 13, 107, 122, 187, 222 chongyun 重韻 196 – 197, 209 – 210 choseong 初聲 34 – 35 coarticulation 108, 131, 140, 211 complementary distribution 25, 27 – 28, 78, 161, 181, 194, 219 consonant cluster 35, 53, 61, 70 deng 等 7 dissimilation 78, 122, 124, 211, 219 distinctive 25, 47 Dueum beopchik 79, 88, 103 eonhae 諺解 19 epenthesis 16, 50 fanqie 反切 xix, 11, 15, 23n12

jongseong 終聲 34 – 35 jungseong 中聲 34 – 35 kaihe 開合 6, 57, 105, 107 kaikou 開口 6 Kan–on 漢音 22, 107

minimal pair 25 neutral tone 31 Optimality Theory 52 outer source layer 16 palatalization 79, 88 phonogram 226 phonotactic constraint 16, 30, 49, 51, 53, 109, 122, 133, 150, 187, 222 – 224 qingchun 輕唇 13

Go–on 吳音 22, 107 gugyeol 口訣 1

rhotacized vowel 28 rhyme dictionary 2 rhyme table 2, 3, 21, 57, 107, 194, 198

Hanyu pinyin xix, 25, 45, 48, 54 – 55 hekou 合口 6 hyangchal 鄉札 1, 3 hypercorrection 103, 110, 124, 178

Sanskrit 21, 156, 206, 207 sheshang 舌上 13, 15, 194 – 195 shetou 舌頭 13, 15, 195 she 摄 107

234  Index Sino-Japanese 10, 12, 22 Sino-Korean i, xxi, 1, 2, 4 – 6, 10 – 14, 16 – 19, 22, 24, 57, 93, 107, 187 – 188, 193, 205, 207, 209, 211 – 212, 219, 225 Sino-Vietnamese 22, 88, 117, 155 – 156, 224 Standard Korean Pronunciation 31 surface form 7, 27 – 28, 50 syllable preference 106, 181 – 182 syllable structure 8 – 9, 24, 29, 35, 44, 53, 205, 224 TBU (Tone Bearing Unit) 29 transliteration 21 – 22, 47, 156, 206 – 208

underlying form 7, 27 – 28, 31, 33 vowel length 35 vowel shift 223 Wade–Giles 45, 54 xieyin 叶音 203 xilian 系聯 194, 208n2 Yale Romanization xix, 33, 56n2 – 3 zero–initial syllable 9 zhengchi 正齒 13 zhisi 支思 rhyme 204 zhongchun 重唇 13